FACEBOOK ADS.

Meta is the only paid channel, and the account is now down to one live ad set: the CMS 2027 Final Rule campaign buying completed webinar registrations inside Florida, at a $40 daily budget. The seven state CMS campaign and the Page Likes campaign were paused at ad set level on Sept 18 to conserve budget, and both Agency Ceiling campaigns closed on Sept 16 on 5 registrations across $255.15. Legacy stays paused. The thing to know about the last few days is spend, not creative: the one live campaign has closed over its $40 budget three days running, at 157, 170 and 130 percent, so the month no longer fits at the nominal rate and the open call is an account level spending limit. Two cautions carry into everything below. Every figure here is a ratio, and a ratio is silent about whether anything ran. And with only one campaign buying registrations there is no second arm to compare against, so day to day swings are noisy and are not read as verdicts on the creative. For the short version of all of this, see the Analytics Snapshot linked at the top.

By the numbers

THE ACCOUNT IS STILL OFF, AND IT STAYED STILL. Oct 11 closed at $0.00 on every campaign, ad set and ad, with no delivery, no impressions and no charge. The activity log carries no owner action after the 9:02pm Sept 22 pause, so there was no restart. Its only lines since Sept 30 are ten custom audiences Meta itself created at 10:55am Oct 8, which are neither an owner action nor delivery, and nothing has been added since. Meta restated nothing this check could see, so the winning ad still reads 85 registrations across $717.98 and the Florida campaign 89 at $8.53 across $759.38, both re-read on the Oct 12 pull. All three campaigns still read ACTIVE because the pauses are at ad set level. WHAT IS LEFT. About $311 of the roughly $500 went unspent in September, today is Oct 12, and nothing is spending it. A restart at the nominal $40 is $40, and at the rate the last seven delivering days ran it is about $45. The worst day on record ran 170 percent of budget, about $68, so even the worst case fits and the spending limit stays a backstop rather than a necessity. The winning ad's frequency read 3.39 for a twentieth morning, 0.11 short of its fatigue threshold with no second creative behind it, and the page no longer prints a crossing date because a paused ad has none. The CMS 2027 angle decayed on Oct 1 and nothing was bought on it. BILLING: the $4.40 Account billed event at 4:07am Sept 30 stays the only charge since $25.00 at 7:39pm on Sept 22, and no new charge or owner action appears since (the log's only lines since 6:16pm Sept 30 are the ten Meta generated custom audiences of 10:55am Oct 8). It is recorded and not explained. AD LIBRARY: 445 ads across all 21 pages, 426 yesterday. The diff printed no revival, and every new ad was checked against ad ids across the whole archive. No tier one page shows a new agent recruiting concept. The tier one advocacy page added two consumer event headlines (a town event and a painting workshop), both on a capped page. A tier three lead vendor swapped all eight of its ads for eight fresh ids with the same two headlines, created Oct 11, which is a refresh and not a disappearance. A tier three agency tooling page added a batch of five untitled ads created Oct 9 to Oct 11. Several complete pages returned ids the archive already held from earlier pulls but not from yesterday, which says the index varies day to day, not that the pages relaunched anything. Five pages returned zero and five hit the 50 cap. Batch size is a proxy for activity only and is not a claim about spend. Nothing was published, paused, edited or budgeted by this check.

Internal ad-account metrics, Mon Oct 12 morning pull. Every ad set is still paused by the owner and Oct 11 closed at $0.00 with no delivery anywhere and no restatement. Today is a partial day read at about 8:30am, so its prices are early and flattered by conversions arriving ahead of the impressions that earned them; the closed-day numbers are directly below, and the charts under them run on complete days except where a bar is marked as still open. Every price and ratio on this page is a measure of how the money performed, so none of them can report that the money stopped. Read the spend figures first.

Yesterday, complete

A NINETEENTH DAY OF NOTHING, AND AGAIN NO RESTATEMENT. Oct 11 closed at $0.00 on every campaign, ad set and ad, with no impressions and no charge. Every ad set has read PAUSED since the owner stopped the last one at 9:02pm on Sept 22, and the activity log carries no owner action since (its only lines are ten custom audiences Meta itself created at 10:55am Oct 8, and nothing after them), so status and delivery agree. The three campaigns and the ad still read ACTIVE, which is the ad set level trap this page has named before. Meta moved no earlier day that this check could see: the winning ad's lifetime still reads 85 registrations across $717.98, and all three campaign lifetimes were re-read unchanged on the Oct 12 pull. WHAT IS LEFT. About $311 of the roughly $500, nothing spending it, and the month it was meant for has ended. The CMS 2027 angle decayed on Oct 1, and nothing was bought on it.

Every ad that spent, Oct 11

Every ad in the account that took any delivery on Oct 11. Ads not listed spent nothing that day.

Registrations, day by day

This chart is the Florida CMS campaign only, on the one creative this page spent a week calling worn out, and since Sept 14 it is no longer the whole CMS story: a second campaign launched that day on a new hook across seven states and has its own chart below. Keeping them apart matters, because averaging two campaigns that buy the same event at different prices produces a number nobody can act on. Read the last four bars together and read the money as carefully as the price. Sept 13 finished at 78 cents with no registration because the ad account was disabled for non-payment. Sept 14 finished at $23.38 with no registration on a full funded day. Sept 15 bought six at $5.66 on $33.93. Sept 16 bought three at $7.73 on $23.18, and that bar is short because delivery across the whole account faded out that day, normal until about 9am and gone by 1pm, not because this campaign was throttled. A series that does this is a spread rather than a trend, and every story this page has tried to lay over it, an audience ceiling, a weekday effect and a wear step, has been falsified within a week of being written. The Agency Ceiling campaigns are kept off this chart and have now closed: 5 registrations across $255.15 is a result rather than a rate. Each pair of bars is one day: how many registrations it bought, beside what each one cost, with the weekday and the date underneath. Daily spend now lives in the Day by day section rather than on the axis, which keeps this chart readable as the run grows. A tall teal bar next to a short gold one is a good day.

The new hook, day by day

The seven state campaign, launched Sept 14 on the Same Day script. This is the fresh creative and the geo re-render in one build, and it is the first time this account has had two campaigns buying the same registration event with different creative, which makes the chart above and the chart here the closest thing to a controlled comparison it has ever run. Two complete days have now run side by side and the older creative has won both: 4 at $8.12 against 6 at $5.66 on Sept 15, then 1 at $11.71 against 3 at $7.73 on Sept 16. Lifetime the two campaigns have gone from 24 cents apart to $1.85 apart, $9.02 here against $7.17 there. Sept 16 is the day worth reading closely rather than discounting. Delivery across the account faded out during the day, so both campaigns were cut short on the same curve from the same budget, which is exactly the shock a paired test absorbs and a time series does not. What it exposed is allocation: in the hours the account did run, Meta gave the Florida campaign $23.18 and this one $11.71 from identical $40 budgets. The campaign is four days old and inside its learning phase, which is a real defence and a shrinking one. The reading that breaks it is a complete day above $12, and Sept 16 came in at $11.71.

Follows, day by day

The same read for the Page Likes campaign, from the Sept 2 relaunch forward rather than the campaign's whole life, because the ads before that relaunch were a different set at a different price. The delivery worry this chart carried for several mornings is closed: the campaign took 110 percent of its budget on Sept 17 and 42 percent by 9am today. What is not closed is the price. Its three dearest days of the whole relaunch are among its last four, and the dashed average line has moved with them. Read the level beside the ratio on the short bars: a $1.01 follow on ten follows is not the same claim as a $0.95 follow on thirty eight, and the chart draws them the same height. What no bar here shows is the thing that actually moves this campaign, which lives one level below it. Five ads were live in this ad set and Meta funded one: Brand Ad took 70 percent of the money on Sept 17 and every cent of Sept 18. This page argued for a week that Meta was starving the set's best creative and that claim was wrong, but the lifetime prices underneath it say something plainer. Brand Ad bought follows at $0.93 over $156.59, the dearest of the seven ads this campaign ever ran. The ad it replaced reached $0.82 over $220.48, on more money than Brand Ad ever had. The owner switched that ad back on at 11:55am on Sept 18, three minutes before pausing the whole ad set, so it now sits ACTIVE inside a paused set: armed, and ready if this campaign restarts. The campaign closed on 748 follows at $0.87 across $653.47.

Does the day of the week explain it?

Closed. Sunday Sept 27 did not deliver at all because every ad set is paused, so it is not a reading and is not added. The record stays at one agreement against six contradictions. The panel closed on Sept 13 and every repeat reading since has been recorded rather than acted on. MONDAY IS THE SIXTH WEEKDAY TO BE READ TWICE AND IT BEHAVES LIKE THE OTHER FIVE. Sept 7 bought four registrations at $8.51; Sept 21 bought two at $19.41, a factor of 2.3 on almost the same money, 97 percent of budget against 85. Nothing about the calendar changed between them. That takes the record to ONE agreement, Tuesday, against SIX contradictions: Wednesday, Friday, Saturday, Thursday, Friday again and now Monday. Every weekday this account has read twice has disagreed with itself except one. A panel that disagrees with itself six times out of seven is measuring noise, and the Thursday pair remains the cleanest demonstration: $19.01 on Sept 10 against $4.98 on Sept 17, a factor of nearly four on the same weekday and almost the same money. Friday is still the only weekday read three times, at $4.06, $9.53 and $12.54, and the third is not comparable on money because Sept 18 spent 157 percent of budget where the other two spent about 95. A panel that was already retired does not get resurrected by a reading it cannot control, and the rows are recorded rather than read. Sept 22 is a Tuesday and it does not join the panel. It bought no registration, so it has no price, and it stopped at 9:02pm when the owner paused the ad set, so it is not a full day of delivery either. The Tuesday row keeps Sept 15 and Sept 8, the one agreement on record. The panel stays on the page as the record of a killed hypothesis rather than a live one. Its value now is that it is the falsification, not the finding.

What the next two weeks should produce

Rebuilt on the delivering days between Sept 10 and Sept 16, and this window is in worse shape than any before it. Sept 13 is excluded because the account was disabled that day and bought nothing anywhere, and this sentence is the label on that filter. Sept 16 is included and it is the reason every central figure below fell: the account delivered for about nine hours that day and spent $86.02 of $209 committed, and dropping it would have propped all three cards up at exactly the moment that would have been most convenient. So these numbers rest on a seven day window holding one dead day and one half delivered day, and the honest expectation is that they are too low rather than too high. Three cards rather than two, because a second CMS campaign exists now, and its card is still a placeholder rather than a forecast, with a range set deliberately at half to one and a half times its central figure. Neither Agency Ceiling campaign is forecast: both stopped on Sept 17 and they closed on 5 registrations across $255.15, which is $51.03 each.

How the calls are holding up

Two ways of holding this account, and the process running it, to account. The decisions table grades recommendations that have already resolved. The projections table holds every dated call open until its date, then writes what actually happened beside what was projected. As the dates below arrive the pending rows fill in, so a week from now this reads as a record of what was expected against what came true rather than a fresh set of confident guesses.

Decisions, graded

28 of 41 graded calls right owner 3 of 5 · this process 25 of 36 The owner's calls held without exception until today, and today one of them came back mixed. The Agency Ceiling campaign was launched into a dated webinar at $43 a day and it finished at $43.47 a registration, six times what the Florida CMS campaign pays. It is graded mixed rather than wrong for a reason worth stating plainly: this page does not know what an agency owner registration is worth to HPM and will not invent a figure to grade a call with. This automated process is behind on its own, and on Sept 17 it added a new kind of error to the record. Its previous failures were all of one shape, reading a pattern into a series too short to carry one. This one was different and worse: it read a tool that had stopped drawing rows as an account that had stopped delivering, published it as fact across four surfaces, and was corrected by the owner rather than by itself. The check that would have caught it costs one pull. Weight everything below accordingly. A third shape arrived the same week and it is the subtlest of the three. For three mornings this page ran what it called a controlled creative test and published a verdict from it. The two arms carried the same claim and the same belief, so the verdict was never available from the data. Nothing in the numbers was wrong; the label on top of them was. It took a cold outside reviewer to ask what the arms were testing, which is a question the content engine already exists to force and which this page never put to its own comparison. The older shape is still the common one: it reads a pattern into a series too short to carry one, then reverses itself. It invented an audience ceiling that was not there. It wanted to hold the budget on Sept 7 when the owner was right to raise it. It moved the re-render off the critical path on a single quiet frequency reading and put it back the next morning. It decided the new campaign's statics did not convert after two funded days, and one of them converted the next day. It downgraded the Page Likes forecast on a five day fall that reversed on the first day delivery resumed. Against that, it was right to insist on verifying the registrations, right to cut the two dead CMS cells, right repeatedly to hold the budget through the dearest days on the account, right to retire the weekday panel a day before its own date, right to refuse to scale the Agency Ceiling campaign on its early price, right to name the licensee campaign as the weaker of the pair, and right for five mornings that a fresh CMS hook was the lever worth pulling. That last entry has now gone through three stages and it is the most instructive thing in the table. On Sept 15 it was right about the action and wrong about the reason. On Sept 16 the reason got worse, because both creatives' click-through doubled on the same day, so the number it argued from was tracking the calendar. Today the seven state campaign has lost both complete days and its lifetime price has moved from 24 cents behind the Florida campaign to $1.85 behind it. The recommendation is now the only part of that argument still standing, and it is standing on the fact that a second campaign exists to compare against rather than on the creative inside it being better. A fourth stage was added on Sept 17: the asset that shipped was never a fresh hook at all. Its locked script carries the same claim and nearly the same payoff line as the ad it replaced, and it was never registered in the content engine, so the lever this table spent five mornings recommending was never actually pulled. A fourth shape arrived on Sept 18 and it is the one that should change how a reader weights everything above. The first three shapes are all errors of inference: a pattern read into too short a series, a tool mistaken for the world, a label put on a comparison it did not fit. This one is an error of measurement, and it is worse because the numbers are real and the arithmetic on them is correct. Twice in a week this page has read a window that begins after the thing it is describing has stopped, and then described the silence. It reported an ad as starved by Meta for four mornings when that ad had been the heaviest funded and cheapest in its set for the five days before the window opened, and had then been switched off. It reported a frequency series reversing when what changed was which entity the number was read from. In both cases the cure is the same and it costs one query: pull the series, not the point, and say which unit the series is in. Read this table as saying: weight its cut recommendations heavily, weight its refusals to act heavily, discount anything it infers from fewer than about five readings per unit, and treat its explanations far more sceptically than its recommendations, because the record now contains a case where the recommendation survived, the mechanism was falsified, the instrument behind the mechanism turned out to be measuring something else, and the thing recommended then lost on the merits.

Projected against actual

Each row is a call with a number and a date. It stays pending until the date arrives, then the actual is written beside the projection. This is the feedback loop: come back after the weekend, after one week, after two, and read what was expected against what happened.

17 of 40 dated calls right 2 still open The calls that set two outcomes against each other come in. The calls that name an absolute threshold come in too, and today the table shows why that is not the same as being useful. The Same Day price call asked whether the new hook holds under $10 a registration, it does at $8.84, and it grades right while the campaign it was defending lost both complete days to a campaign buying at $6.35 over the same two days. A bar the loser clears comfortably is a bar in the wrong place, and the fix is to write the comparison into the call rather than the level. The pattern that keeps repeating is graded-right-for-the-wrong-reason, and it has now happened three days running. The Agency Ceiling outage call predicted the campaign would finish on four registrations because the outage would take its days; it got every day back, spent $73.58 across them, bought nothing, and finished on four. The frequency restatement landed one hundredth outside its range having been carried there by an outage and a half delivered day rather than by the decelerating climb it argued for. A prediction that survives the destruction of its own premise is a coincidence and is recorded as one. SEPT 22 IS THE COUNTER-EXAMPLE AND IT IS WORTH AS MUCH AS THE FAILURES. The instrument assertion written on Sept 21 named a number, a date and a bar, cleared the number at $38.67 against $48, and cleared its reason too, because the thing it asked about, a fourth consecutive day over budget, did not happen. It is the second half of a matched pair: the same instrument graded WRONG on Sept 19 and RIGHT on Sept 22, and the pair says something neither row says alone, that the overspend is real and is not every day. That is what the instrument assertions were adopted for, and it is the first time one of them has earned its keep in both directions. The re-dated rows are closed and the record on them is clean. Both were re-dated when the outage removed their evidence, both moved earlier rather than later, and both closed against the argument attached to them. The Service Fire row is closed, and how it closed matters more than that it did. It was carried for three mornings as a call waiting on somebody to fund an ad. The ad was paused the whole time, so the row could never have resolved, and the question it asked had already been answered in data this page had not pulled: that ad spent $90.57 across five days and bought follows at $0.83, cheaper than the ad Meta funds instead. It is graded wrong rather than closed clean, because a call whose premise is false is a failure of the check that wrote it. The hardest thing on this table is the row that had to be graded twice. A call written specifically to test whether this page's own frequency instrument means anything was graded right in the small hours of Sept 18 and regraded wrong the same morning, because the reading it was graded against turned out to be taken from a different entity than the rest of its series. That is the best designed row here and it still confirmed a fault instead of catching one, for a reason that generalises: the grader and the faulty input were the same morning's work. An instrument check that runs inside the instrument is not independent. What caught it was a physical impossibility in the number, a reach jump larger than the ad can reach in a day, and that kind of sanity bound is worth more here than another dated call. Today's two new rows are both instrument assertions rather than market forecasts, which is the only class this process is now allowed to open on its own, and both are written so that they can only convict this page rather than excuse it. SEPT 23 GRADED THE TWO OLDEST OPEN FORECASTS AND BOTH ARE WRONG. The fourteen day registration and follow forecasts written on Sept 9, the last market forecasts of the retired kind, closed at roughly half of what they projected. Both are graded on the days that existed, and the days the owner's pauses removed are declared on each row rather than used as cover, because neither gap could have been closed by them. They are the strongest argument on this table for the Sept 17 decision to stop writing that class of call. Sept 24 graded the paused-account assertion right and did not replace it. No new row was written, because a daily repeat of the same zero check adds a line without adding a test, and the morning check reads delivery against status every run whether or not a row asks it to. Sept 25 to Oct 2 had nothing due and wrote nothing, for the same reason.

Day by day

Every day since the registration campaigns went live, so the account reads as a trend rather than a single morning. Read the spend bar first on this page rather than last: on Sept 16 it is a third of the day before it, because delivery across the whole account faded out during that day, and every price on that row still reads normally. Each row carries its weekday, kept as a label rather than as an explanation: the weekday panel below was closed on Sept 13 and three of the four weekdays read twice have now contradicted their own first readings. The registration count is every registration the account bought that day, across all campaigns, not the CMS campaigns alone. The registration price column is deliberately blank from Sept 14, when a second CMS campaign launched at a different price, because one number across two campaigns buying the same event at different prices is not a price anyone can act on. The per-campaign prices are in the two charts below.

1. WEDSep 2
   $20.480 registrationsnone5 follows$1.17 each

   Launch day. The registration campaigns went live late and barely delivered.

2. THUSep 3
   $121.584 registrations$21.05 each23 follows$1.63 each

   Full delivery across three campaigns, and only four registrations for it.

3. FRISep 4
   $106.5310 registrations$6.00 each54 follows$0.86 each

   The turn. All ten on CMS at $4.06. Legacy paused mid day, Page Likes expanded to seven states.

4. SATSep 5
   $37.526 registrations$4.47 each18 follows$0.60 each

   Both live campaigns underspent and still produced the cheapest finished day.

5. SUNSep 6
   $91.285 registrations$8.97 each52 follows$0.89 each

   The heaviest day on the account. Both campaigns ran well past budget and the registration price doubled.

6. MONSep 7
   $77.364 registrations$8.51 each62 follows$0.70 each

   The first finished day on the raised budgets. CMS stayed under its new $36 and still paid $8.51; Page Likes went over and bought its cheapest follows yet.

7. TUESep 8
   $83.9610 registrations$3.88 each54 follows$0.84 each

   The best finished day on the account. Ten registrations at $3.88 on $38.79 of CMS spend, which settled the budget question in favour of the raise.

8. WEDSep 9
   $85.109 registrations$4.65 each53 follows$0.82 each

   The first finished day at the raised budgets. CMS ran to $41.83 and still bought nine registrations at $4.65, so two of the three cheapest days on the account are now the two most recent.

9. THUSep 10
   $76.992 registrations$19.01 each51 follows$0.76 each

   The break in the run. CMS spent close to its budget and bought two registrations at $19.01, more than four times the day before. Page Likes did not notice, and finished at $0.76.

10. FRISep 11
    $107.924 registrations$9.53 each47 follows$0.75 each

    The first day all three campaigns delivered, $107.92 against $119 committed. CMS came in at $9.53, between Thursday's $19.01 and the $4 run before it, and the Agency Ceiling campaign bought its first registration at $34.44.

11. SATSep 12
    $109.504 registrations$9.24 each36 follows$0.61 each

    The Agency Ceiling campaign outspent both established ones, $50.75 against a $43 budget, and bought three registrations at $16.92. CMS came in at $9.24, its second day near $9.40. Page Likes spent $21.81 of $36.

12. SUNSep 13
    $1.540 registrationsnone0 followsnone

    The account was disabled for non-payment. Three card charges failed on this day, delivery stopped in the morning, and the finished total is $1.54 against $119 committed with no registration and no follow anywhere on the account.

13. MONSep 14
    $120.372 registrationsnone37 follows$1.02 each

    The account came back and the shape of it changed. The balance was paid, delivery resumed, and two new campaigns launched: a seven state CMS campaign on a new hook and a second Agency Ceiling campaign on a licensee audience, taking committed spend from $119 to $209. Both registrations came from the new CMS campaign. The two established registration campaigns spent $53.88 between them and bought nothing.

14. TUESep 15
    $181.5810 registrationsnone38 follows$0.95 each

    The best day the account has finished. $181.58 against $209 committed, 87 percent, with every campaign delivering and 10 registrations bought. Both CMS campaigns converted, the Florida one at $5.66 and the seven state one at $8.12, and the Florida campaign beat the seven state one on the first complete day they have both run. The two Agency Ceiling campaigns took $79.06 of the money and bought nothing for the third complete day running.

15. WEDSep 16
    $86.025 registrationsnone10 follows$1.01 each

    Delivery decayed from about 9am and had stopped entirely by early afternoon. $86.02 against $209 committed, 41 percent, and $77.63 of that was already spent by 8:30am, so the whole of the rest of the day added $8.39 across five campaigns. Hour by hour in the ad account time zone, the three live campaigns ran normally to 9am, then in pennies, and their last delivering hours were 11am on the seven state campaign, 1pm in Florida and 2pm on Page Likes. The day before, the Florida campaign delivered until 11pm. Nothing in the account's own reporting says why, no charge failed and no campaign showed a payment error. Two of the five campaigns were the Agency Ceiling pair, which the owner switched off early. Every price on the day still reads normally, which is the part worth carrying forward.

16. THUSep 17
    $108.9810 registrationsnone33 follows$1.20 each

    Closed at $108.98 against the $116 now committed, 94 percent, and it settles the question Sept 16 opened: that day was an event, not a new level. Ten registrations, eight of them to the Florida campaign at $4.98, its best complete day since Sept 8. The seven state campaign bought two at $14.80, and its A arm was switched off by the owner during the evening. Page Likes took $39.51 for 33 follows at $1.20, dearer than its band, with Brand Ad taking 70 percent of the money.

17. FRISep 18
    $117.297 registrations$13.21 each19 follows$1.31 each

    The last row that mixes three campaigns, and the most misleading total in the series unless it is read with its denominator. $117.29 against the $116 committed at midnight, which reads as a normal 101 percent day; but two of the three ad sets were paused at 11:58am, so most of this was bought by one campaign running 157 percent of its own budget. Florida alone took $62.71 of a $40 budget across a full twenty four hours. From here the account commits $40 a day rather than $116.

18. SATSep 19
    $68.078 registrations$8.51 each0 followsnone

    The first complete day of the one ad set account, and it closed at $68.07 against $40 committed, 170 percent. That is higher than Sept 18's 157 percent, so the overspend is not a one-off. Eight registrations at $8.51 on the money, which is the best output of the three overspending days.

19. SUNSep 20
    $51.911 registration$51.91 each0 followsnone

    Third consecutive day over budget, $51.91 against $40, 130 percent. One registration on the money, which is the dearest priced day this account has recorded, and the count can still rise under seven day click attribution because the day closed only hours ago. Read the level beside the ratio: one registration prices nothing reliably.

20. MONSep 21
    $38.822 registrations$19.41 each0 followsnone

    97 percent of the $40 committed, the first closed day inside budget after three at 157, 170 and 130 percent. Two registrations at $19.34. The twenty four hourly rows summed to exactly $38.67 and 754 impressions on the Sept 22 pull; Meta restated the day to $38.82 and 755 on the Sept 24 pull, ordinary backfill recorded rather than smoothed over.

21. TUESep 22
    $29.650 registrationsnone0 followsnone

    74 percent of the $40 committed, and the day stopped at 9:02pm when the owner paused the last live ad set. No registration on the money, a funded day with no result, and the count can still rise under seven day click attribution. The twenty two hourly rows summed to exactly $29.60 on the Sept 23 pull; Meta restated it to $29.65 and 577 impressions on Sept 24.

22. WEDSep 23
    $0.000 registrationsnone0 followsnone

    No delivery anywhere, the first full day with every ad set paused. The activity log attributes the pause to the owner at 9:02pm on Sept 22 and records no action since, so zero is the expected reading and not a failure.

23. THUSep 24
    $0.000 registrationsnone0 followsnone

    The second full day with every ad set paused. No delivery, no charge, no owner action in the log, and no restatement of any earlier day.

24. FRISep 25
    $0.000 registrationsnone0 followsnone

    The third full day with every ad set paused. No delivery, no charge, no owner action in the log, and no restatement of any earlier day.

25. SATSep 26
    $0.000 registrationsnone0 followsnone

    The fourth full day with every ad set paused. No delivery, no charge, no owner action in the log, and no restatement of any earlier day.

26. SUNSep 27
    $0.000 registrationsnone0 followsnone

    The fifth full day with every ad set paused. No delivery, no charge, no owner action in the log, and no restatement of any earlier day.

27. MONSep 28
    $0.000 registrationsnone0 followsnone

    The sixth full day with every ad set paused. No delivery, no charge, no owner action in the log, and no restatement of any earlier day.

28. TUESep 29
    $0.000 registrationsnone0 followsnone

    The seventh full day with every ad set paused. No delivery, no charge, no owner action in the log, and no restatement of any earlier day.

29. WEDSep 30
    $0.000 registrationsnone0 followsnone

    The eighth full day with every ad set paused. No delivery, no owner action in the log, and no restatement of any earlier day. A Meta generated custom audience was created at 6:16pm; it is not delivery.

30. THUOct 1
    $0.000 registrationsnone0 followsnone

    The ninth full day with every ad set paused. No delivery, no owner action in the log, no new charge, and no restatement of any earlier day.

31. FRIOct 2
    $0.000 registrationsnone0 followsnone

    The tenth full day with every ad set paused. No delivery, no owner action in the log, no new charge, and no restatement of any earlier day.

32. SATOct 3
    $0.000 registrationsnone0 followsnone

    The eleventh full day with every ad set paused. No delivery, no owner action in the log, no new charge, and no restatement of any earlier day.

33. SUNOct 4
    $0.000 registrationsnone0 followsnone

    The twelfth full day with every ad set paused. No delivery, no owner action in the log, no new charge, and no restatement of any earlier day.

34. MONOct 5
    $0.000 registrationsnone0 followsnone

    The thirteenth full day with every ad set paused. No delivery, no owner action in the log, no new charge, and no restatement of any earlier day.

35. TUEOct 6
    $0.000 registrationsnone0 followsnone

    The fourteenth full day with every ad set paused. No delivery, no owner action in the log, no new charge, and no restatement of any earlier day that this check could see.

36. WEDOct 7
    $0.000 registrationsnone0 followsnone

    The fifteenth full day with every ad set paused. No delivery, no owner action in the log, no new charge, and no restatement of any earlier day that this check could see.

37. THUOct 8
    $0.000 registrationsnone0 followsnone

    The sixteenth full day with every ad set paused. No delivery, no owner action in the log, no new charge, and no restatement of any earlier day that this check could see. The log's only lines are ten custom audiences Meta itself created at 10:55am.

38. FRIOct 9
    $0.000 registrationsnone0 followsnone

    The seventeenth full day with every ad set paused. No delivery, no owner action in the log, no new charge, no new log line of any kind, and no restatement of any earlier day that this check could see.

39. SATOct 10
    $0.000 registrationsnone0 followsnone

    The eighteenth full day with every ad set paused. No delivery, no owner action in the log, no new charge, no new log line of any kind, and no restatement of any earlier day that this check could see.

40. SUNOct 11
    $0.000 registrationsnone0 followsnone

    The nineteenth full day with every ad set paused. No delivery, no owner action in the log, no new charge, no new log line of any kind, and no restatement of any earlier day that this check could see.

41. MONOct 12so far
    $0.000 registrationsnone0 followsnone

    Open day, read at the morning run. Zero because every ad set is still paused and the activity log shows no restart and no new charge.

Campaigns are collapsed to their headline. Open one to see its creative, and click any ad for the full detail.

OffCMS 2027 Final Rule, seven statesCompleted registrationsPAUSED at AD SET level, midday Sep 18. The campaign still reads ACTIVE at $40 / day0 live2 paused

Yesterday $29.74 for two registrations at $14.87 before the ad set was paused at midday. Its last dayToday so far No delivery. $0.00

OffCMS 2027 Final Rule, FloridaCompleted registrationsPAUSED at AD SET level by the owner, 9:02pm Sep 22. The campaign and its ad still read ACTIVE at $40 / day0 live4 paused

Yesterday $0.00 and no delivery. Oct 11 was the nineteenth full day with its ad set pausedToday so far No delivery. $0.00

OffPage LikesPage followsPAUSED at AD SET level, midday Sep 18. The campaign still reads ACTIVE at $36 / day0 live15 paused+8

Yesterday $24.84 for 19 follows at $1.31 before the ad set was paused at midday. Its last dayToday so far No delivery. $0.00

OffThe Agency CeilingCompleted registrationsOff since the owner stopped it on Sep 160 live4 paused+1

Yesterday No delivery. $0.00Today so far No delivery. $0.00

OffLegacy ProgramCompleted registrationsOff since Sep 40 live4 paused

Yesterday No deliveryToday so far No delivery

The benchStaged, not in a live campaign yetNot assigned12 staged+5

The same ads again, grouped by where each one stands in Meta right now. 0 live, 0 in rotation, 0 pending, 11 upcoming, 30 paused.

Live0

Published and delivering now

In Rotation0

Established, still running

Pending0

Submitted, in Meta review

Upcoming11

Built or staged, not published

Hybrid / sell anywhereRendered 09-02

Hybrid Social A · Windshield

Social motion, cinematic look, voiced. The week that disappears behind the windshield; $250 in leads on the table.

Hybrid webinar4:5 + 9:16 staged

Hybrid / sell anywhereRendered 09-02

Hybrid Social B · Same Commission

Opens on the payoff: same commission, less driving. The hook A/B.

Hybrid webinar4:5 + 9:16 staged

Legacy / your bookRendered 09-02

Legacy Social A · The Day You Stop

Social motion, clean Gala look, voiced. What happens to your book the day you stop; Cancun stay on the table.

Legacy webinar4:5 + 9:16 staged

Legacy / your bookRendered 09-02

Legacy Social B · Who Gets It

Twenty years building it, who gets it? Three ways out, none of them back to the carrier.

Legacy webinar4:5 + 9:16 staged

Agency builderRendered 09-02

Agency Social A · Leadership Decision

Social motion, cinematic look, voiced. Not a money decision; the phone at eleven on a Saturday. Co-hosted event on the table.

Start an Agency webinar4:5 + 9:16 staged

Agency builderRendered 09-02

Agency Social B · Producer to Owner

Twenty years of overnights; the honest picture before the boat.

Start an Agency webinar4:5 + 9:16 staged

Legacy / your bookRendered 09-02

Releases Social A · Lose Your Renewals?

Social motion, cyberpunk look, voiced. Renewals follow your tax ID; free commission audit on the table.

Navigating Releases webinar4:5 + 9:16 staged

Legacy / your bookRendered 09-02

Releases Social B · Less Trapped

Opens on the answer: most agents are far less trapped than they think. How a release actually works.

Navigating Releases webinar4:5 + 9:16 staged

Page Likes / followsRendered 09-02

Brand · Thirteen Years

General branding, clean Gala, silent. Since 2013, A+ with the BBB, 950+ agents, 42 states, as service.

Brand / retargeting pool4:5 + 9:16 staged

Page Likes / followsRendered 09-02

Brand · You Lead, We Handle the Rest

General branding, clean Gala, silent. The back office as the hero.

Brand / retargeting pool4:5 + 9:16 staged

Page Likes / followsRendered 09-02

Brand · Keep What You Build

General branding, clean Gala, silent. The Three Uniques.

Brand / retargeting pool4:5 + 9:16 staged

Paused30

Off, held, or consolidated out

Legacy / your bookPre 08-25

Legacy Image Ad

“You built a book.” The cost-per-view winner of the Legacy set.

Legacy webinarProduced the account's first Meta registrations; off with the Legacy campaign, paused Sep 4 at $35.52 per registration against the CMS campaign's price on similar money. The comparison offered against that pause has now closed and it closed on the pause's side: both Agency Ceiling campaigns finished dearer than Legacy, $43.47 on the broad one and $81.27 on the licensee one

2 registrations at $29.50 lifetime

Legacy / your bookExisting

Legacy Video Ad (9:16)

“Go see the world.” Legacy motion for Stories and Reels.

Legacy webinarCampaign off, and it never earned delivery against the image

23 impressions and no views on its last week

Agency builderPublished 09-10

Campaign closed Sept 17Took 99 then 98 percent of its campaign and bought nothing on eitherEnds at 3 registrations for $121.92, a price of $40.64Its licensee copy bought the buy's cheapest registration, $6.58

AC · Snippet 42s

The 42 second cut. The agency principal who cannot get off the service desk long enough to sell.

The Agency Ceiling webinarClosed with its campaign on Sept 17. It was the ad Meta funded first and it carried the buy early, one registration on Sept 11 at $12.80 and two on Sept 12 at $10.95 each, prices inside what the CMS campaign was paying that week. Then it stopped. Across every funded day after Sept 12 it bought nothing, including 99 percent of Sept 15's money and 98 percent of Sept 16's, and it ends at $121.92 for three registrations, a price of $40.64. The verdict on the ad is inseparable from the verdict on the allocation: it was given almost the entire campaign and could not convert on it, while its two ad set siblings were never given enough to fail properly. Its licensee copy did the one good thing in the whole buy, a registration at $6.58 on the last delivering day

campaign closed. 3 registrations at $40.64 lifetime, all of them bought on or before Sept 12

Agency builderPublished 09-10

Campaign closed Sept 17. No registration anywhere, everA $206.94 CPM on its last day, the dearest thousand on recordStarved in the broad campaign, funded in the licensee one, failed in both

AC · Promo 2:52

The long promo. What the ceiling actually is, and the three fires behind it.

The Agency Ceiling webinarClosed with its campaign having never bought a registration in either of them. It is the one Agency Ceiling ad that got a fair hearing and failed it: starved in the broad campaign, where Meta gave it 48 cents on the campaign's last delivering day, and funded in the licensee one, where it took $14.70 on Sept 15 and $7.45 on Sept 16 and produced nothing from either. Its last day carried a $206.94 CPM on 36 impressions, the most expensive thousand impressions anything on this account has ever bought. Nothing about this cut has earned a funded day in a future campaign

campaign closed. No registration in either campaign across its whole life

Agency builderPublished 09-10

Campaign closed Sept 17Lifetime spend never moved after Sept 12. Starved, not beatenIts one registration came on its third day, at $15.94Its licensee copy ended with one link click on 66 impressions

AC · Offer card

The offer card static, straight to the register page.

The Agency Ceiling webinarClosed with its campaign, and it ends as the clearest case on this account of an ad that was starved rather than tested. It is the ad that falsified this page once already, written off after two funded days and then converting on its third at $15.94. The dated call asking whether that was a one-off resolved today and grades wrong, on an ad whose lifetime spend in the broad campaign never moved after Sept 12 while AC | Snippet 42s ran on to $121.92. Its licensee copy did get money, $6.79 on the campaign's last day for one link click on 66 impressions, and bought nothing across its whole life. So the broad copy is unproven and the licensee copy is beaten, and those are different findings that a single name on a board would have merged

campaign closed. 1 registration, bought Sept 12, and no broad delivery after it

Agency builder

Agency builderPublished 09-10

Switched off Sept 13 after $15.63 and no registration. The cut graded right

AC · You Lead

You lead, we handle the rest. The back office as the answer to the ceiling.

The Agency Ceiling webinarThe second static, switched off by the owner after three funded days as the campaign's most expensive silence. The timing looked uncomfortable at the time, because the static it most resembles had converted the day before. The campaign has now closed and the final prices settle it: nothing was bought anywhere in the buy after Sept 12, the ad it resembled never converted again, and the 42 second cut ran on to $121.92 and bought nothing more. The cut removed an ad from a set that had stopped working rather than one that was about to start

switched off Sept 13 after $15.63 and no registration, and the cut graded right

CMS 2027 rule

CMS 2027 ruleLocked 09-14

Switched off by the owner at ad level, evening of Sept 17Ends on 8 registrations at $9.63 lifetime across $77.02Lost both complete paired days, $8.12 then $11.71Not a creative verdict: both arms carried the same claim

Same Day A · Know It Cold

The Judgment Arc rebuild, written off Florida. You know the 48-hour rule cold; from October 1 the wait is gone, same scope, same day.

CMS 2027 webinarThe seven state campaign, two complete days in, beaten on both. It was briefed as a fresh hook and reported that way for three days; corrected Sept 17, because its locked script carries the same claim as the ad it is measured against and it was never registered in the content engine, so this pair cannot be read as a creative test. Sept 15 it bought 4 at $8.12 against the Florida ad's 6 at $5.66; Sept 16 it bought 1 at $11.35 against 3 at $7.73. Its lifetime price has gone $7.32, $7.72, $8.43 while the ad it was built to replace went $7.34, $7.48, $6.97. The Sept 16 result comes with the strongest caveat and the weakest excuse available: delivery across the account faded out during the day, so this ad was cut short, and so was the ad beating it, on the same budget and the same curve. What that day did expose is allocation. From identical $40 budgets Meta gave the Florida ad $23.18 and this one $11.35. It is still four days old and still inside its learning phase, which is a real defence and a shrinking one. Its life ends at $77.02 for 8 registrations, $9.63 each, against the Florida ad's $7.18. Two things should not be concluded from that. It is not a verdict on copy, because both arms carry the same claim. And it is not a clean verdict on geography either, because this ad never left its learning phase and was cut four days in

switched off by the owner on the evening of Sept 17. Ends on 8 registrations at $9.63 across $77.02

CMS 2027 rule

CMS 2027 ruleLocked 09-14

PAUSED with its ad set at midday Sept 18Ends on 3 registrations across $36.81 and never produced a priceTook all of its ad set's last day, $29.74 for 2 at $14.87

Same Day B · Ready Client

The B hook of the same script, opening on the lost client rather than on recognition. Still losing a ready client to a two-day scope wait?

CMS 2027 webinarThe B hook of the locked Same Day script, and the question on it has started to answer itself in the encouraging direction. It went live on Sept 16 and took 3 percent of its ad set's money that day, which is the concentration pattern the dated call was written to catch. This morning it has taken $3.36 of the ad set's $16.85, which is 20 percent. Meta is splitting rather than parking it. Nothing about its performance is readable yet: 31 impressions, three link clicks and a 9.68% rate that is a rounding artefact of a tiny sample rather than a result, and no registration. It ends without ever having produced a price worth quoting. Its last day is the only one that showed anything and what it showed is not about the ad: with Same Day A gone it took the whole ad set, $29.74 for two registrations at $14.87, which is what removing an incumbent from an ad set does rather than a result. Three registrations across its life does not price a creative, and the concentration call written against it graded WRONG on its mechanism even though its number cleared

PAUSED with its ad set at midday Sept 18. Ends on 3 registrations across $36.81

CMS 2027 ruleBuilt 08-31

Switched off, no registration in nineteen days

CMS V1 · 48-Hour Wait

The 48-hour wait is dead. Five CMS 2027 changes agents need to know.

CMS 2027 webinarPlate video cell in the dedicated CMS campaign; a 0.88% link click-through against the winner's 2.78% lifetime. The cut recommended on Sep 8 has been made, though the change is not attributed here because the account activity log was unreachable

switched off after $4.76 and no registration in eleven days

CMS 2027 ruleBuilt 08-31

Switched off, no registration in nineteen days

CMS V2 · Still Waiting

Compliant rewrite of V1, a deliberate variant to test the hook.

CMS 2027 webinarPlate video cell in the dedicated CMS campaign; a 0.64% link click-through, the lowest on the account. Switched off alongside V1

switched off after $1.54 and no registration in eleven days

Hybrid / sell anywhereBuilt 08-31

Hybrid · Sell Anywhere

Done with windshield time. Sell from anywhere, keep the book.

Hybrid webinarCampaign off after two days, so the test is unfinished, not failed

2 views at $0.42

Page Likes / followsSince 08-25

$2.14 and no follow yet this morning133 follows at $0.88 lifetime, against a campaign $0.86No delivery on Sept 16

Image B · Ownership

The follow winner. Meta concentrated the Interests budget here.

Page followsThe withdrawn cut recommendation still reads better than the cut would have: $0.88 lifetime on $117.12 and 988 people reached, against a campaign $0.87, and cheaper than the ad taking the whole budget. It took $2.21 yesterday on 74 impressions and bought no follow, and nothing at all today. Its engagement ranking reads Above average on the account's own scoring, which still sits oddly against the Below average this page recorded earlier and is still worth a second look rather than a restatement

133 follows at $0.88 lifetime on $117.12, $2.21 yesterday and nothing today

Page Likes / followsBuilt 08-31

Page Likes 9:16 Motion

“Grow your book. Keep what you build.” Reels-native motion.

Page followsOff since the Sep 4 pull, alongside Grow Your Book. The Page Likes set ran five live ads, not seven, and as of midday Sept 18 all five are paused with the ad set

switched off after 13 follows at $1.05 over the week

Page Likes / followsBuilt 08-31

Switched back ON by the owner at 11:55am Sept 18, three minutes before its ad set was pausedACTIVE with effective_status ADSET_PAUSED. Armed, cannot deliver268 follows at $0.82 on $220.48, the most money and the second cheapest price in the set$90.57 across Sept 8 to Sept 12 at $0.83, so it was never starvedCheaper over its life than the ad now taking the whole budget

Service Fire

Back-office support angle. Builds the retargeting pool.

Page followsEverything this page published about this ad was wrong, and it was wrong in both directions at once. It was called starved by Meta for four mornings, then called untested, then called active as recently as yesterday. It is paused, it stopped delivering after Sept 12, and before it stopped it was the best funded ad in the campaign: $16.17, $18.39, $23.96, $23.13 and $8.92 across Sept 8 to Sept 12, buying 109 follows at $0.83, and $220.48 over its life for 268 follows at $0.82. The $1.24 quoted against it for three mornings is its final $3.72 on the way down. The error was mechanical rather than interpretive: the page read the current day and a trailing window that starts after the ad had stopped, and never once pulled its day by day series, which took a single query and is what settled it. What stands now is a comparison, not a claim: this ad bought follows at $0.82 on real money, and the ad that took the whole budget buys them at $0.93 on less. THE OWNER ACTED ON THAT. The activity log shows him switching this ad back to Active at 11:55am on Sept 18, three minutes before he paused the whole Page Likes ad set at 11:58am to conserve budget. So it reads status ACTIVE with effective_status ADSET_PAUSED: armed, zero spend on its first nominal day because the set above it is off, and it will deliver the moment the campaign restarts with nothing further to do. This page recorded the recommendation as WITHDRAWN yesterday, which was wrong, and the correction is on the decisions row as well as here. It is also a small lesson about the activity log: the log does not carry ad-level PAUSES, which had led this page to treat it as unreliable for ad status, but it does carry ad-level ACTIVATIONS, and reading it is what caught this

Switched back ON by the owner at 11:55am Sept 18, then its ad set was paused at 11:58am. ACTIVE, ADSET_PAUSED, armed and unable to deliver. 268 follows at $0.82 on $220.48

Page Likes / follows

Page Likes / followsSince 08-25

Image C · Belonging

Page follows. Belonging angle.

Page followsPaused in the Sep 2 relaunch; it never earned delivery

5 follows at $1.03 lifetime, starved throughout, campaign ad set off

Page Likes / follows

Page Likes / followsSince 08-25

Took 97 percent of the money on Sept 16 and all ten follows10 follows at $0.97 on a campaign that stopped delivering by 2pm79 percent of the money this morning at $2.63Lifetime price drifted from $0.65 to $0.87 as it scaled

Brand Ad

Page follows. Brand angle.

Page followsIt is where the money goes: 70 percent of the campaign yesterday and every cent of it so far today. Its lifetime price has drifted from $0.65 to $0.93 as it scaled, which is the standing observation on this card and remains the useful one, an ad's price on scraps is not its price on a budget. What changed today is what it should be compared against. Its $0.93 over $156.59 is the DEAREST lifetime price of the seven ads this campaign has run, and the comparison is like for like rather than scraps against scale: Service Fire reached $0.82 over $220.48, more money than this ad has ever had. Every other live ad in the set is cheaper too, at $0.70, $0.74, $0.83 and $0.88. Meta has concentrated the campaign on its most expensive creative, which is the opposite of the story this page told for a week, and it is a stronger finding because it rests on lifetime prices at comparable scale rather than on single days

PAUSED with its ad set at midday Sept 18. 168 follows at $0.93 lifetime on $156.59, and all 19 of the last day's follows

Page Likes / followsBuilt 08-27

Page Likes 1:1 Motion

Page follows video, in the Broad ad set.

Page followsBroad ad set is off; never truly ran

CMS 2027 ruleRendered 09-02

PAUSED with its ad set by the owner, 9:02pm Sept 2285 of the account's 105 registrations, at $8.45 lifetimeSept 22: $29.60 and no registration before the pauseReads status ACTIVE, effective_status ADSET_PAUSEDStill the only ad in its ad set, nothing to rotate toFrequency frozen at 3.39, 0.11 below the 3.5 threshold

CMS Social A · The Clock

Social motion, cyberpunk look, voiced. The 48-hour wait is ending October 1; five changes, one move to make.

CMS 2027 webinarThe account's winning ad, and as of 9:02pm on Sept 22 it cannot deliver: the owner paused its ad set, so it reads status ACTIVE with effective_status ADSET_PAUSED and took nothing after the 21:00 hour. Its last day, Sept 22, spent $29.60 and bought no registration on 11 link clicks over 575 impressions, 1.91 percent, which is one funded day with an open attribution window and no control on the account, so it is recorded rather than read as wear. Before it, three closed days ran over budget at 157, 170 and 130 percent and one inside it at 97. Ad level lifetime is 85 registrations at $8.45 across $717.98 on 11,759 impressions, after Meta restated its last two days by twenty cents on Sept 24. Frequency froze at 3.39 on a reach of 3,470 on the Maximum window, 0.11 below the 3.5 fatigue threshold. THE THING A RESTART BRINGS BACK: this ad set still holds no second creative, so if the owner switches it back on it crosses its fatigue threshold inside a day or two with nothing to rotate to.

85 registrations at $8.45 lifetime at ad level across $717.98. Paused with its ad set by the owner at 9:02pm Sept 22

CMS 2027 ruleRendered 09-02

Switched off Sept 10 after 4 registrations at $8.78

CMS Social B · Still Waiting

Same beats as A, opens on the question. The hook A/B for the social motion format.

CMS 2027 webinarIt converted on scraps early and never converted again once Meta funded it: three straight days of its largest delivery with nothing on any of them, and a lifetime price that drifted from $7.56 to $8.78. The owner switched it off rather than waiting for the Sept 13 date this process had put on it, which was the right call. All three of the campaign's dearest days have landed since, with no second ad for Meta to move money to

switched off after 4 registrations at $8.78 lifetime

Page Likes / followsRendered 09-02

Page Likes · Gala Motion

“Grow your book. Keep what you build.” in the clean Gala motion look, a look-test against the running 9:16.

Page followsThe cut recommended on Sep 4; with two of its three cells already gone, the Gala look test cannot produce a clean read

switched off after 9 follows at $1.57, the most expensive live follow

Page Likes / followsPublished 09-06

No delivery on Sept 16 or Sept 1765 follows at $0.70, the cheapest in the set

Page Likes · 30-Day Release (Caption)

Leave anytime, keep your book. The 30-day release as motion, with the caption written out.

Page followsThe cheapest follow price in the set over its life, 68 at $0.70 on $47.48. It took $2.07 yesterday and bought three at $0.69, and it has taken nothing at all today. The caveat that belongs on it is real: $47.48 is a third of what the funded ad has had, so this price has not been tested at scale and could drift upward the way Brand Ad's did, from $0.65 to $0.93. That caveat does not apply to Service Fire, which reached $0.82 on more money than any ad here, and keeping the two straight is the point. This ad is a promising untested price; that one is a proven one

68 follows at $0.70 lifetime, 3 of Sept 17's follows on $2.07, and nothing today

Page Likes / followsPublished 09-06

35 cents on Sept 16, its only money in three days32 follows at $0.74 lifetime

Page Likes · 30-Day Release (No Caption)

The same release cut with the primary text left empty, so the film carries it alone.

Page followsTen cents yesterday on three impressions and nothing today. On $23.69 of lifetime spend this ad has the thinnest record in the set and $0.74 is closer to a rumour than a price. It is listed so the board shows what is actually live rather than what is actually running: five ads are live in this ad set and Meta funds one of them

32 follows at $0.74 lifetime on $23.69, ten cents yesterday and nothing today

Page Likes / followsPublished 09-06

No delivery on Sept 16, after its first readable day on Sept 15

Page Likes · Back Office (Caption)

The back office as the hero. What an agent stops doing once the service work is handled.

Page followsNow paused, which this page had not recorded. Its whole life is $3.13 of spend and four follows, so $0.78 is a number rather than a read and nothing is lost by it being off. It runs the corrected render with the on screen text reading SELL: EMPTY

PAUSED at ad level by the owner on the evening of Sept 17. 4 follows at $0.78 on $3.13 lifetime

Page Likes / followsPublished 09-06

Still barely any delivery, nothing on Sept 16 or Sept 17

Page Likes · Back Office (No Caption)

The same back office cut with no primary text.

Page followsIt took $7.63 yesterday and bought seven at $1.09, its first real day, which moved its lifetime from $0.54 on scraps to $0.83 on $10.84. That move is the standing lesson of this campaign in miniature: a price built on three dollars goes up when real money arrives. It has taken nothing today. It runs the corrected render

13 follows at $0.83 lifetime on $10.84, and 7 of Sept 17's follows at $1.09

Page Likes / followsRendered 09-02

Page Likes · Service Fire Motion

Motion cut of the running Service Fire static, Gala look, silent.

Page followsOne of the four Sept 2 experiments the owner cut to lower cost per follow

switched off Sept 3 at $1.73 per follow

Page Likes / followsRendered 09-02

Page Likes · Three Ways Out Motion

Motion cut of the staged Three Ways Out static, Gala look, silent.

Page followsOne of the four Sept 2 experiments the owner cut

switched off Sept 3 at $2.99 per follow, the weakest of the six

Agency builderStaged 09-01

Build the Agency

Agency principal to Start an Agency. Round-two hook.

Page followsOne of the four Sept 2 experiments the owner cut; the agency-builder angle

switched off Sept 3 at $1.41 per follow

Legacy / your bookRendered 09-01

Your Book Has Three Ways Out

The three Legacy exit paths, one image.

Legacy webinarNever earned delivery against the image ad, and now off with the Legacy campaign pause

$0.70 lifetime, starved throughout

Legacy / your bookStaged 09-01

AEP, Without Doing It Alone

Support through AEP. Round-two hook.

Page followsOne of the four Sept 2 experiments the owner cut; the support-through-AEP angle

switched off Sept 3 at $1.90 per follow, on a 0.79% click-through rate

Legacy / your bookStaged 09-01

Never got a fair read, 8.00% CTR on its one funded day

Keep Your Book (30-Day Release)

Leave anytime, keep your book. The 30-day release.

Legacy webinarOut-clicked the image winner the day before the Legacy pause and never got a fair read; the strongest reason to bring Legacy back

$10.70 lifetime, an 8.00% click-through rate on its one funded day

By category

Content engine

Brief queue

Corpus review

What we found out

Ad test creatives

New creative built for testing lives on its own page, so this one stays a read of what is actually running.

Timeline

1. Aug 25First four Meta ad sets launched: Legacy v1/v2/v3 and Page Likes Interests.
2. Aug 27Registration tracking fixed across all five webinar configs; Legacy links moved to the fbclid-safe register URLs; Google and Microsoft switched off, Meta stays live.
3. Aug 31Legacy consolidated to one funded ad set (v1 at $30 a day, v2 and v3 paused); CMS V1, CMS V2, and Hybrid built and published; two new Page Likes ads shipped.
4. Sep 1Round two creative library staged (four new hooks). Morning pull: Legacy 773 views at $0.14, Page Likes 139 follows at $0.83, frequency 1.5 to 1.7, still zero attributed registrations.
5. Sep 2The objective switch. The Legacy traffic buy is retired and two registration campaigns go live on the fixed pixel, optimizing for completed sign-ups instead of landing page views. Budgets settle at about $30 a day per group, and the round two statics and social motion hooks all reach the auction by that night.
6. Sep 3The registration column moves off zero and the first four hold up in WebinarJam. The owner prunes four Page Likes experiments that had pushed cost per follow up, and switches the older Brand Ad back on.
7. Sep 4The day the account changed shape. CMS Social A starts converting and the owner verifies the registrations against the WebinarJam list. Legacy is paused on the evidence, 2 registrations at $33.15 against CMS at $8.14 on nearly identical spend, and its budget moves to CMS, a deliberate 10 percent step that did not reset the learning phase. Page Likes expands from Florida to seven southeastern states, which CMS cannot copy because its winning ad names Florida three ways.
8. Sep 5Nothing was changed and the campaign got cheaper anyway. Both live campaigns underspent their budgets and still produced the cheapest finished day so far. Grow Your Book and the 9:16 motion cut go off, leaving Page Likes on three ads.
9. Sep 6The same pattern, plus the first real constraint. The winning ad's frequency climbs from 1.62 to 1.83 in a day against a reach of 825 in one state, so the geo re-render moves from a nice-to-have to the item with a clock on it. CMS Social B converts again on almost no delivery. Image B Ownership drops to a Below average engagement ranking. The account activity log was unreachable, so nothing here is attributed to a person. Later the same morning the owner raised both live campaigns 10 percent on that evidence, CMS from $33 to $36 and Page Likes from $30 to $33, taking committed spend from $63 a day to $69. Both campaigns were read back Active after the write, so the silent pause that followed a budget change on Sep 4 did not recur.
10. Sep 7The first day the numbers argue against the pending budget increase. Yesterday the CMS campaign ran to $44.86 on a $33 budget, roughly the spend the increase was asking for, and bought five registrations at $8.97 where the smaller day before it bought six at $4.47. The winner's frequency reached 2.01 on a reach of 1,098, still inside Florida. Four wave two Page Likes motion cuts cleared review and are delivering, the Brand Ad is switched off, and the account activity log was unreachable for a third day, so nothing here is attributed to a person.
11. Sep 8The first finished day on the raised budget, and it did not settle the question. CMS spent $34.03 of its new $36 and bought four registrations at $8.51, a second expensive day after Sep 6's $8.97, on less money rather than more. Then this morning opened at four registrations for $15.08, $3.77 each, the cheapest start the account has recorded, and the lifetime price improved from $6.88 to $6.72 as attribution caught up. The winner's frequency reached 2.17 on a reach of 1,269, still inside Florida. Page Likes bought 62 follows at $0.70, its cheapest finished day, and the Brand Ad is switched back on and delivering. The account activity log was unreachable for a fourth day, so nothing here is attributed to a person.
12. Sep 9The judgement date, answered twice over. Sept 8 finished at $38.79 of CMS spend and bought ten registrations at $3.88, the best finished day the account has recorded on both volume and price, which breaks the tie the two expensive days created. Across five finished days the cost per registration does not track spend, so the $8.97 and $8.51 run reads as a wobble rather than a ceiling. The owner also reconciled the registrant list against WebinarJam and the registrations are real, which closes a caveat that had been carried for four mornings. On that pair of answers both budgets were raised another ten percent, CMS from $36 to $40 and Page Likes from $33 to $36, taking committed spend from $69 a day to $76; both campaigns were read back Active after the write, the fourth clean budget change in a row. CMS V1 and CMS V2 are switched off, leaving the ad set on its two working hooks, and the cut recommended three mornings running on Image B Ownership is withdrawn because at $0.89 against a campaign $0.85 it is at par. The winner's frequency reached 2.30 on a reach of 1,428. The geo re-render is deliberately on hold: the creative does not exist yet, so the account is buying volume now and will meet the frequency threshold without a wider audience ready.
13. Sep 10The first finished day at the raised budgets came in and it held: $41.83 of CMS spend, nine registrations at $4.65, the second cheapest day on the account, which puts two of the three cheapest days in the last two days. Page Likes bought 53 follows at $0.82, a seventh straight day inside its band. Wednesday finished and joined the cheap weekday group, so the split is now four cheap days against two dear ones with no overlap. The reading that changes the plan is the frequency curve: the winner went 2.30 to 2.37, an increment of seven hundredths against 0.13, 0.16, 0.18 and 0.21 before it, while its reach grew 296 in a day, the largest single-day gain of the run. The climb is decelerating, so the Sept 17 threshold date is revised to about Sept 21 and the geo re-render moves from critical path to insurance. What is tightening instead is the winner's click-through, 3.14% lifetime against 4.11% in week one, a 23.6 percent fall that is close to the plan's 25 percent rule. CMS Social B took its two largest days of delivery and produced nothing, moving its lifetime price to $8.69. Nothing was published, paused, edited or budgeted, and the Meta reporting connector failed authentication for a sixth day, so nothing here is attributed to a person.
14. Sep 11The account grew a third campaign and the CMS half got thinner in the same twenty four hours. The Agency Ceiling campaign, built and switched on by the owner, started delivering this morning: $43 a day, four ads in one broad ad set across seven southeastern states, a hard stop on Sept 17 when its webinar runs, and $7.07 spent with no registration yet. Committed spend goes from $76 a day to $119. Against that, Sept 10 finished as the CMS campaign's worst day since its learning phase, $38.01 for two registrations at $19.01, and the owner switched CMS Social B off, which leaves the ad set on a single creative. That creative tripped the plan's click-through fatigue rule this morning at 3.00% against a 4.11% first week, a 27.0 percent fall, which grades yesterday's dated call right a day early. Frequency contradicted yesterday rather than confirming it: the increment came in at 0.16 after the 0.07 that had looked like deceleration, so the threshold sits near Sept 19 and the step is noisy rather than bending. Page Likes did none of this, 51 follows at $0.76. Nothing was published, paused, edited or budgeted by this check, and the Meta reporting connector failed authentication for a seventh day, so nothing here is attributed to a person.
15. Sep 12The day the weekday story broke and the new campaign answered its question. Friday finished at $9.53 a registration on $38.10, which is neither Thursday's $19.01 nor a return to the $4 run before it, so the outlier call this page made yesterday is graded wrong and the middle outcome is the informative one. Friday is also the first weekday this panel has read twice: $4.06 on Sept 4 against $9.53 on Sept 11, on almost identical money, a gap wider than the one the panel was built to explain. Sorted by date the two dearest complete days on the account are the two most recent, which is the shape of a creative wearing rather than a calendar. The Agency Ceiling campaign bought its first registration on Sept 11 at $34.44, two days ahead of its dated call, with one of its four ads buying it and the other three spending $21.64 between them for nothing. A method correction goes with all of it: the click-through rule tripped on Sept 11 against a cumulative lifetime average, which falls by construction once current days sit below the historical mean, so the daily rate is the honest instrument and it has been flat near 1.9% for three days at about half the first week. Nothing was published, paused, edited or budgeted by this check, and the Meta reporting connector failed authentication for an eighth day, so nothing here is attributed to a person.
16. Sep 13Saturday answered the discriminating call and the account grew a second problem while it did. CMS finished at $9.24 a registration on $36.94, above the $7 line, so the trend reading beat the weekday reading and the panel is retired for good: Saturday is the second weekday to be read twice and, like Friday, it moved from the cheap group to the dear one on similar money. The three dearest of the campaign's nine complete days are its three most recent, and the last two sit two cents apart, which is the shape of a step rather than a slide. The Agency Ceiling campaign had its second funded day and tripled its output: three registrations at $16.92, taking it to four at $21.48 lifetime, and one of them was bought by the static offer card that this page had recorded the previous morning as part of a pair that bought nothing. The owner switched AC | You Lead off after three funded days and $15.63 with no registration, the only ad-level change on the account since Sept 10. The finding nobody was looking for is in the steady campaign: Page Likes follows have fallen on five consecutive days, 62, 54, 53, 51, 47 and then 36, on spend that went from $43.33 to $21.81 against an unchanged $36 budget, while the cost per follow stayed inside the band it has held for ten days. A ratio held while the thing under it shrank. Nothing was published, paused, edited or budgeted by this check, and the Meta reporting connector failed authentication for a ninth day, so nothing here is attributed to a person.
17. Sep 14The account stopped. Meta attempted to charge the card ending 5235 three times on Sept 13, for $18.00, $2.33 and $2.33, and all three failed, the day after six charges of $18.00 on the same card all cleared. The ad account is disabled, every live campaign shows Payment error in place of a delivery status, and the balance due is $19.10. Sept 13 finished at $1.54 of spend against $119 committed with no registration and no follow anywhere on the account, and Sept 14 has bought nothing at all. Two dated calls this page was holding open, on whether the CMS price had stepped to a new level and on whether Page Likes had a delivery problem, named Sept 13 and Sept 14 as their evidence; both are re-dated rather than graded, because the evidence was removed rather than contradicted. The winning ad's frequency fell for the first time in the series, 2.65 to 2.57, and the fall is an artefact twice over: the ad served 14 impressions in three days, and two report windows covering identical delivery returned 2.65 on a reach of 2,185 and 2.57 on a reach of 2,258 on this one morning. Nothing was published, paused, edited or budgeted by this check, no payment was made or attempted by it, and the Meta reporting connector failed authentication for a tenth day.
18. Sep 15The account came back and arrived with two more campaigns than it left with. The balance was paid, every campaign reads Active, the current balance is $8.78 and no charge has failed since Sept 13, though the billing threshold has moved off $18 to $21 and $10 amounts. Delivery resumed on Sept 14 and the day finished at $120.37 against $209 committed. The shape change is the owner's: the Same Day script was locked on Sept 14 and shipped as a new campaign, HPM | CMS 2027 Final Rule | Leads | SE7 at $40 a day running Same Day A | Know It Cold across a broad seven state ad set, recorded at the time as the fresh hook and the geo re-render in one build, and corrected on Sept 17 to the geo re-render only, because its locked script carries the same claim as the ad it replaced; and a second Agency Ceiling campaign at $50 a day runs the same ads against a licensee audience. Committed spend goes from $119 a day to $209. The new campaign bought both of the day's registrations, 2 at $5.13, while the Florida ad spent $23.38 and bought none, its first funded day without a registration, on a day its click-through read 2.10% and five people reached the landing page. The two Agency Ceiling campaigns spent $49.09 between them and bought nothing, which takes the broad one's lifetime price from $21.48 to $34.81 on a count frozen at 4. Page Likes spent 105 percent of its budget, which answers the delivery question opened on Sept 13 in the direction of a dip, and cost $1.02 a follow doing it, the first time it has been over a dollar, because Brand Ad took $33.75 of $37.65 at $0.99 while Service Fire took nine impressions. Nothing was published, paused, edited or budgeted by this check, no payment was made or attempted by it, and the Meta reporting connector failed authentication for an eleventh day.
19. Sep 16The best finished day the account has had, and the morning its own reasoning came apart. Sept 15 spent $181.58 against $209 committed, 87 percent, with every live campaign delivering and 10 registrations bought. The two CMS campaigns ran side by side on identical $40 budgets for the first complete day and the Florida one won both ways: 6 registrations at $5.66 against the seven state campaign's 4 at $8.12, and 3.73% link click-through against 3.30%. This entry originally called that a creative result and it is not one; both arms carry the same claim, and the correction was made on Sept 17. The click-through reading is the one that matters beyond this account. This page argued for five mornings that the winner was worn because its day level click-through had settled near half its first week, in a band of 1.83% to 2.15% across five readings. Yesterday it read 3.73% and the new creative read 3.30%, both roughly double the day before, and this morning they read 3.87% and 3.13%. Two different creatives in two different geographies moved together, so the band was measuring the day and not the wear. Page Likes came back under a dollar at $0.95 as Brand Ad's share fell from 90 percent to 72, and Service Fire, the ad this page has called starved rather than tested, was given $3.72 and bought follows at $1.24, dearer than the Brand Ad it was being compared against. The two Agency Ceiling campaigns took $79.06 and bought nothing for a third complete day; the licensee one then bought its first registration this morning at $5.53 at ad level, and both stop tomorrow. The two unpublished drafts are cleared and Same Day B | Ready Client is live on $1.63, so the locked A/B has both arms for the first time. Nothing was published, paused, edited or budgeted by this check, no payment was made or attempted by it, and the Meta reporting connector failed authentication for a twelfth day.
20. Sep 17Delivery across the account faded out during Sept 16 and the day finished at $86.02 against $209 committed, 41 percent, having spent $77.63 of it before 8:30am. The remaining fifteen and a half hours added $8.39 across five campaigns. The hourly breakdown in the ad account time zone shows all three live campaigns delivering normally to about 9am, then in pennies, and stopping completely between 11am and 2pm; the Florida campaign ran to 1pm and its thirteen hourly rows sum to exactly its day, against a Sept 15 that delivered until 11pm. That shared curve across three campaigns makes it an account level event rather than a campaign level one. The first version of this entry said delivery stopped dead at 9am; that was read off a grid that renders only nine rows until its own scrollbar is dragged, and it was corrected the same day. Nothing in Ads Manager explains it: every campaign read Active, no charge failed, there is no payment error, the balance is $10.80 and the threshold has taken $33.00 and two $15.00 charges since. Delivery is normal again this morning, $52.06 by 8:25am against the $116 now committed. Every price on the short day read normally, which is the finding: a page built out of ratios had nothing to say about whether the account ran. Inside it the paired comparison held, because both arms were truncated at the same hour on the same budget: the Florida ad bought 3 at $7.73 against the seven state ad's 1 at $11.71, and it took $23.18 against $11.71 from identical $40 budgets. Lifetime the two have moved from 24 cents apart to $1.85 apart. Both Agency Ceiling campaigns stopped on schedule and the buy closes at $255.15 for 5 registrations, $51.03 each, taking $93 a day off committed spend. Same Day B took 20 percent of its ad set this morning against 3 percent yesterday. Nothing was published, paused, edited or budgeted by this check, no payment was made or attempted by it, and the Meta reporting connector failed authentication for a thirteenth day.
21. Sep 18Sept 17 closed normally and settles the question Sept 16 opened: $108.98 against $116 committed, 94 percent, delivered across the whole day rather than front-loaded. Sept 16 was an event, not a regime, and the right response to it was to do nothing, which is what was done. Ten registrations, eight to the Florida campaign at $4.98 on its full budget, its cheapest complete day since Sept 8 and its third win in a row over the seven state campaign. The day inside the seven state campaign is the one that mattered: $29.60 for two registrations at $14.80, split $24.16 to Same Day A for one and $5.44 to Same Day B for one, and the owner switched Same Day A off that evening on exactly that arithmetic. It is the first ad level cut on this account made by the owner rather than proposed by this page. The campaign stays live at $40 a day on Same Day B alone, so committed spend is unchanged at $116, and whether to pull it is an open owner decision. Two dated calls resolved. The three day comparison graded RIGHT, all three days to Florida. The click-through call graded WRONG: the two CMS ads closed 1.16 points apart against a 1.0 bar, so that number carries both a day term and an ad term and this page read the sum as the second one for five mornings. This entry also carried two claims that were corrected later the same day and the corrections are recorded here rather than quietly removed. It said the frequency runway had reversed, from 2.91 to 2.65 on a reach up 765 to 3,329: that figure was read at campaign level and appended to a series built from the winning ad, the ad itself went 2.91 to 2.97 on a reach of 2,564 to 2,681, and the crossing date is reinstated at about Sept 26. It also said Service Fire took nothing for a fourth day, implying an active ad being denied delivery: the ad is PAUSED, and across Sept 8 to Sept 12 it had been the set's heaviest funded and cheapest ad at $90.57 for 109 follows at $0.83. AT MIDDAY THE ACCOUNT CHANGED SHAPE AGAIN, and this one is the owner's: he paused `CMS 2027 | SE7 | Broad` and `Page Likes | FL | Interests` to conserve budget, with about $500 left to the end of the month. He paused them AT AD SET LEVEL, so both campaigns and all their ads still read ACTIVE while neither can spend a cent, which is the third distinct way in one week that something here has read as on while not running. Committed spend falls from $116 a day to $40 and `CMS 2027 | FL | Leads` is the only live ad set left. Both campaigns were recorded here that day as closing on 11 registrations at $10.31 across $113.38 and 746 follows at $0.88 across $652.94; those were partial-day reads taken while the day was still running, and the true closing figures, pulled on Sept 19, are 11 at $10.35 across $113.83 and 748 follows at $0.87 across $653.47. Two dated rows close as DESTROYED EVIDENCE rather than grade, both declared: the Sept 20 seven state comparison, and this morning's own delivery assertion, whose denominator was halved four hours after it was written. The Service Fire recommendation is WITHDRAWN, because its ad set is now off. Nothing was published, paused, edited or budgeted by this check; both account changes today were the owner's.
22. Sep 19Sept 18 closed and the number that matters is not the account total. The account finished at $117.29 against the $116 committed at midnight, which reads as an ordinary 101 percent day and is nothing of the sort: two of the three ad sets were paused at 11:58am, and the Florida campaign ran all twenty four hours to $62.71 on a $40 daily budget, 157 percent, by far the largest day it has ever spent and well outside the roughly 25 percent a day Meta documents. It is real spend and not a rendering fault, which was checked three ways: the twenty four hourly rows sum to exactly $62.71 with every hour present, impressions and reach rose with the money at 956 and 724 against a series running near 650 and 520, and the activity log carries no budget change. Nor is it one campaign's property. Through 11:59am, the hour the other two were paused, Florida had taken 78 percent of its budget, the seven state campaign 74 percent of its own and Page Likes 69 percent of its own, so all three were pacing between 140 and 157 percent on the same day, which makes it a property of the day. What the extra money bought was worse: 48 percent more impressions than Sept 17, link click-through DOWN from 2.79 to 2.51 percent, cost per link click UP from $2.22 to $2.61, and five registrations at $12.54 where the cheaper day before bought eight at $4.98. That reading is uncontrolled and is labelled so, because the second CMS arm that could have said whether Sept 18 was simply a dearer day was paused at midday; losing the control is a real cost of the pause and it is the first time this page has had to say so. The consequence is arithmetic and it is the day's decision: twelve days at the nominal $40 is $480 against about $500, and twelve days at Sept 18's rate is closer to $750, so the $480 is a floor rather than an estimate. A throttle to $36 scales the floor and not the multiple. The recommendation moves from throttle-or-limit to the limit, and it is a billing action the owner takes. One dated call graded and it graded RIGHT on both bits: the instrument assertion written yesterday to test whether the frequency unit correction held said the series would read above 2.97 on a reach above 2,681 at AD level, and it reads 3.09 on 2,894, with the reach step of 188 well inside a day in which the ad reached 724 people. The crossing moves three days earlier to about Sept 23, and that is the budget rather than the audience. Yesterday's own delivery assertion closes as DESTROYED EVIDENCE and is deliberately not banked although its number cleared: the day passed $98 for a reason opposite to everything it asserted. One thing published here yesterday is corrected. The Service Fire recommendation was recorded as WITHDRAWN; the activity log shows the owner ACTED ON IT, switching the ad back to Active at 11:55am and pausing its ad set at 11:58am, so it is armed rather than moot and will run the moment the campaign restarts. The Ad Library archive produced its first real diff against a like-for-like basis and there is no material change: 179 ads across nine tier-1 pages, one new competitor ad which is a third copy of a concept that page already ran twice. The diff's own NEW CONCEPTS line was checked against ad ids and discarded as an artefact of this archive's title transcription rather than a market event. Nothing was published, paused, edited or budgeted by this check.
23. Sep 20The Sept 19 instrument assertion resolved and it resolved WRONG, which is the whole finding. It said Sept 19 would close at or under $48, 120 percent of the $40 committed, and that anything above meant the overspend was not a one-off. Sept 19 closed at $68.07, 170 percent, a HIGHER multiple than Sept 18's 157. Then Sept 20 closed at $51.84, 130 percent, a third consecutive day over budget. Three days over budget is a pattern, not a wobble. The benign explanation was checked and killed in the same pass: the parked catch-up idea was that Meta was recovering a trailing weekly average, but the trailing seven days spent $302.98 against the $280 that seven days at $40 commits, which is 108 percent of the weekly budget and not a recovery toward it. What the three days bought splits sharply and is uncontrolled either way: Sept 19 bought eight registrations at $8.51 with link click-through back up at 3.23 percent, Sept 20 bought one at $51.84 at 2.21 percent, the dearest priced day the account has recorded. One registration is not a price, the count can still rise under seven day attribution, and with the seven state arm off there is no second unit to run the lockstep check against, so none of this is read as a verdict on the creative. The money is the decision and it is shorter than it was: Sept 19 and Sept 20 took about $120 of the roughly $500 the owner had, leaving about $380 for ten days, and ten days at even the nominal $40 is $400. The recommendation is unchanged in direction and firmer in degree: an account level spending limit is now the only instrument that fits, because a throttle scales the nominal number while the overspend scales the multiple. It is a billing action and the owner's alone. No run happened on Sept 20, so this entry and the Sept 21 pull cover both days. Nothing was published, paused, edited or budgeted by this check.
24. Sep 21Two runs on one date: a midnight run covered the closed days Sept 19 and Sept 20, and this morning run re-verified them and read the open day. Sept 20 re-verifies EXACTLY unchanged at $51.84 and one registration, with no attribution rise, and its twenty four hourly rows sum to the day total to the cent. The new reading is that Sept 21 is pacing below all three overspend days, and the cut that shows it is the same hours rather than the budget: through hour 08 it has spent $14.03 against $23.86 on Sept 19 and $16.44 on Sept 20, which on those days' own close-to-hour-08 multiples of 2.85 and 3.15 puts it near $40 to $44. That is a leading read on the assertion resolving tomorrow, not a result, and it does not reopen the settled finding that the overspend is a regime. Billing is clean with the threshold at $17, three charges on Sept 20 and one at 6:12am on Sept 21, none failed, and the activity log carries no owner action and no budget change since Sept 20, so the overspend is not an edit. The frequency panel found a fault in itself: Sept 21 read 3.21 at 00:46 and 3.26 at 09:20, so the sampling hour is worth 0.045 while a real day is worth 0.06 to 0.11, and the series had been carrying an undeclared clock-time index. The morning reading is the convention-consistent one and the crossing moves to about Sept 24 with a two day band. No second Ad Library pull was taken, because the archive is one file per day and an eight hour diff on a capped moving window manufactures exactly the false signals the Sept 21 appearance guard was built to suppress. Nothing was published, paused, edited or budgeted by this check.
25. Sep 22The overspend stopped at three days. Sept 21 closed at $38.67, 97 percent of its $40 budget, and the instrument assertion written against it grades RIGHT on both bits; its 24 hourly rows sum to exactly $38.67. The trailing week still reads 114 percent of the weekly budget, so the spending-limit recommendation stands. Frequency 3.35 on 3,400, and the crossing held at about Sept 24 for the first time in ten readings. The Ad Library corrected itself: a tier-1 competitor page returned the IDENTICAL seven ad ids it was reported to have lost, so the Sept 21 disappearance was an empty result from the index rather than a page going quiet, and a zero-coverage guard was added to the diff tool. Next: grade the Sept 22 closed day against the $48 bar, and watch the winner cross 3.5 with no second creative behind it.
26. Sep 23The owner paused the last live ad set, CMS 2027 | FL | Leads, at 9:02pm on Sept 22, per the activity log, so the account has no delivery anywhere while all three campaigns still read ACTIVE. Sept 22 closed at $29.60 with no registration, and its instrument assertion graded RIGHT on both bits; the pause removed at most $9.48 of possible spend and did not decide the grade. The two fourteen day forecasts written on Sept 9 graded WRONG, 52 registrations against a low end of 69 and 324 follows against 698. Frequency froze at 3.39. One tier-1 competitor page built 45 ads across eleven concepts in one afternoon, four of them revivals. CORRECTED SEPT 24: those four were not revivals. Every one had been running throughout; the page refreshed them with new ad ids. Nothing was published, paused, edited or budgeted by this check.
27. Sep 24Sept 23 closed at $0.00 on every unit on the account, the first full day with every ad set paused, and the instrument assertion that a paused account stays at zero graded RIGHT; the activity log shows no owner action since 9:02pm Sept 22. Meta restated Sept 21 and Sept 22 upward by twenty cents in all, so the Florida lifetime reads $759.38. Frequency read 3.39 again. Yesterday's report of four competitor revivals is withdrawn: checked against ad ids, the four concepts had been live in every pull since Sept 19, so the diff was reading a gap in creation dates as a gap in running, and it now separates a refresh from a revival. Nothing was published, paused, edited or budgeted by this check.
28. Sep 25Sept 24 closed at $0.00 on every unit, the second full day with every ad set paused, with no charge, no owner action in the log and, unlike the day before, no restatement, so the Florida lifetime still reads $759.38. Frequency read 3.39 for a third morning and the page stopped printing a crossing date, because two flat steps take the checker to its undated branch. The Ad Library showed no material change. Nothing was published, paused, edited or budgeted by this check.
29. Sep 26Sept 25 closed at $0.00 on every unit, the third full day with every ad set paused, with no charge, no owner action in the log and no restatement. Frequency read 3.39 for a fourth morning and the page still prints no crossing date. In the Ad Library, a small tier-1 recruiting page went from two ads to six with one new concept, and a larger one built a third refresh batch of twelve. Nothing was published, paused, edited or budgeted by this check.
30. Sep 27Sept 26 closed at $0.00 on every unit, the fourth full day with every ad set paused, with no charge, no owner action in the log and no restatement. Frequency read 3.39 for a fifth morning. With four days left, even the worst day on record fits the money, so the spending limit becomes a backstop. The Ad Library showed no material change; four AEP season ads one page dropped are confirmed gone on a second complete pull. Nothing was published, paused, edited or budgeted by this check.
31. Sep 28Sept 27 closed at $0.00 on every unit, the fifth full day with every ad set paused, with no charge, no owner action in the log and no restatement. Frequency read 3.39 for a sixth morning. Three days are left and every rate on record fits the money. In the Ad Library one tier-1 page refreshed its seven ad Spanish language concept with fresh ids on a complete pull, a refresh and not a revival. Nothing was published, paused, edited or budgeted by this check.
32. Sep 29Sept 28 closed at $0.00 on every unit, the sixth full day with every ad set paused, with no charge, no owner action in the log and no restatement. Frequency read 3.39 for a seventh morning. Two days are left and every rate on record fits the money. Nothing was published, paused, edited or budgeted by this check.
33. Sep 30Sept 29 closed at $0.00 on every unit, the seventh full day with every ad set paused, with no owner action in the log and no restatement. The log does carry one Account billed event of $4.40 at 4:07am Sept 30 with no spend behind it, the first charge in eight days, recorded and not explained. Frequency read 3.39 for an eighth morning. Today is the last day of the month with about $311 of the roughly $500 unspent. Nothing was published, paused, edited or budgeted by this check.
34. Oct 1Sept 30 closed at $0.00 on every unit, the eighth full day with every ad set paused, with no owner action in the log and no restatement. The only new log line is a Meta generated custom audience at 6:16pm, which is not delivery. Frequency read 3.39 for a ninth morning. September ended with about $311 of the roughly $500 unspent. Nothing was published, paused, edited or budgeted by this check.
35. Oct 2Oct 1 closed at $0.00 on every unit, the ninth full day with every ad set paused, with no owner action in the log, no new charge and no restatement. Frequency read 3.39 for a tenth morning. Nothing was published, paused, edited or budgeted by this check.
36. Oct 3Oct 2 closed at $0.00 on every unit, the tenth full day with every ad set paused, with no owner action in the log, no new charge and no restatement. Frequency read 3.39 for an eleventh morning. Nothing was published, paused, edited or budgeted by this check.
37. Oct 4Oct 3 closed at $0.00 on every unit, the eleventh full day with every ad set paused, with no owner action in the log, no new charge and no restatement. Frequency read 3.39 for a twelfth morning. Nothing was published, paused, edited or budgeted by this check.
38. Oct 5Oct 4 closed at $0.00 on every unit, the twelfth full day with every ad set paused, with no owner action in the log, no new charge and no restatement. Frequency read 3.39 for a thirteenth morning. The snapshot headline, direction line and restart tile still carried late September wording (three days left, spend past Sept 30) and were corrected. Nothing was published, paused, edited or budgeted by this check.
39. Oct 6Oct 5 closed at $0.00 on every unit, the thirteenth full day with every ad set paused, with no owner action in the log, no new charge and no restatement. Frequency read 3.39 for a fourteenth morning. The forecast card still priced three days to Sept 30 and was rewritten as one change, and a stale Oct 1 phrase on the weekday panel was corrected. Nothing was published, paused, edited or budgeted by this check.
40. Oct 7Oct 6 closed at $0.00 on every unit, the fourteenth full day with every ad set paused, with no owner action in the log, no new charge and no restatement that this check could see. Frequency read 3.39 for a fifteenth morning. Nothing was published, paused, edited or budgeted by this check.
41. Oct 8Oct 7 closed at $0.00 on every unit, the fifteenth full day with every ad set paused, with no owner action in the log, no new charge and no restatement that this check could see; the three campaign lifetimes were re-read and are unchanged. Frequency read 3.39 for a sixteenth morning. Nothing was published, paused, edited or budgeted by this check.
42. Oct 9Oct 8 closed at $0.00 on every unit, the sixteenth full day with every ad set paused, with no owner action in the log, no new charge and no restatement that this check could see; the log's only lines were ten custom audiences Meta itself created. Frequency read 3.39 for a seventeenth morning. Nothing was published, paused, edited or budgeted by this check.
43. Oct 12Oct 11 closed at $0.00 on every unit, the nineteenth full day with every ad set paused, with no owner action in the log, no new charge, no new log line of any kind and no restatement that this check could see. Frequency read 3.39 for a twentieth morning. Nothing was published, paused, edited or budgeted by this check.
44. Oct 11Oct 10 closed at $0.00 on every unit, the eighteenth full day with every ad set paused, with no owner action in the log, no new charge, no new log line of any kind and no restatement that this check could see. Frequency read 3.39 for a nineteenth morning. Nothing was published, paused, edited or budgeted by this check.
45. Oct 10Oct 9 closed at $0.00 on every unit, the seventeenth full day with every ad set paused, with no owner action in the log, no new charge, no new log line of any kind and no restatement that this check could see. Frequency read 3.39 for an eighteenth morning. Nothing was published, paused, edited or budgeted by this check.
46. NextThe account is off by the owner's hand and about $311 remains. The only decision is whether a restart is still wanted now that the CMS 2027 angle has decayed on Oct 1. A restart brings the one creative back as the live question, since it sits 0.11 below its fatigue threshold with nothing behind it. If it restarts, a spending limit at about $310 (an owner only billing action) caps the total whatever the daily rate does. The morning check reads delivery against status every run and will say so the day anything restarts.
47. Oct 1The CMS 2027 angle decayed here. That deadline was the reason spending into a narrowing Florida audience looked like a reasonable trade: the volume was worth more before this date than the audience is worth after it.