The budget moved to the worst ad set. That is the story this week.
Thursday 23 July 2026. Scope: the 12 Sail-named campaigns inside ad account act_305550522185383 (Hotel 32 32). Client-hotel campaigns in the same account were not read and not touched. Nothing was activated, no budget or bid was changed, and no running ad was edited. Three new creative concepts below are proposals only - no ads were uploaded.
1. What happened
On 11 July, Sail Lead Magnet dropped from about $100/day to about $10/day, and ABM went from $0 to roughly $20-50/day. Both campaign daily budgets now read $10. Over the last 7 days ABM took 63% of all Sail spend at $136 a lead, while the main campaign, which converts at $36 a lead, is throttled to about $10/day. That single allocation is worth more than any creative in this brief.
Second thing worth knowing: the ten AI-built variants are now live. As of this morning [01] through [10] are ACTIVE in all six delivering Lead Magnet ad sets and in the ABM ad set. They have taken $1.74 of spend and 133 impressions between them - nowhere near enough to read. The playbook line about "every AI variant has $0 lifetime spend" is now out of date; the new problem is not activation, it is that a $10/day campaign has no budget to distribute across 17 ad names.
2. Sail campaigns - last 28 days
| Campaign | Spend | CTR | Leads | CPL |
|---|---|---|---|---|
| Sail Lead Magnet | $1,651.71 | 1.17% | 45 | $36.70 |
| Sail Lead Magnet - ABM | $408.24 | 1.02% | 3 | $136.08 |
| Sail SBB | $27.36 | 2.18% | 0 | - |
| Total | $2,087.31 | 1.17% | 48 | $43.49 |
By ad, 28 days, everything above $9 of spend:
| Ad | Campaign | Spend | CTR | CPC | Leads | CPL | Freq |
|---|---|---|---|---|---|---|---|
| Video 1 - Gradient | Lead Magnet | $947.29 | 1.16% | $0.15 | 21 | $45.11 | 3.76 |
| Video 2 - Not Gradient | Lead Magnet | $634.99 | 1.17% | $0.17 | 21 | $30.24 | 4.25 |
| Video 1 - Gradient | ABM | $214.72 | 1.34% | $4.13 | 2 | $107.36 | 9.13 |
| Video 2 - Not Gradient | ABM | $100.11 | 0.59% | $7.70 | 0 | - | 6.14 |
| Static Image 3 | Lead Magnet | $57.58 | 1.19% | $0.56 | 2 | $28.79 | 4.03 |
| Static Image 1 | ABM | $40.46 | 0.82% | $5.78 | 0 | - | 4.71 |
| Static Image 2 | ABM | $39.86 | 1.26% | $3.99 | 1 | $39.86 | 4.29 |
| Merged Ads | SBB | $27.36 | 2.18% | $1.24 | 0 | - | 1.15 |
| Static Image 2 | Lead Magnet | $9.14 | 1.00% | $0.48 | 0 | - | 2.00 |
3. Last 7 days
| Campaign | Spend | Share | CTR | Leads | CPL |
|---|---|---|---|---|---|
| Sail Lead Magnet - ABM | $136.29 | 63% | 0.71% | 1 | $136.29 |
| Sail Lead Magnet | $72.21 | 33% | 1.30% | 2 | $36.10 |
| Sail SBB | $8.04 | 4% | 2.73% | 0 | - |
Total $216.54, 3 leads, $72.18 a lead - double the 28-day rate, entirely because of where the money is sitting.
4. Flagged
ABM runs on one ad set against the eblast list, roughly 1,000 people, advantage audience off by design. 28-day frequency on Video 1 there is 9.13; CPC is $4.13 against $0.15 for the same creative in the main campaign. CTR on that ad fell 1.34% (28d) to 0.97% (7d); Video 2 fell 0.59% to 0.49% and has produced zero leads on $100. Adding creative to this ad set reaches nobody new - the list is the ceiling, not the ads.
Inside the main campaign, the Sail Lead Magnet - eblast audience (A+A off) ad set has Video 1, Video 2, Static Image 1 and Static Image 3 sitting at WITH_ISSUES with Meta error 4469003, "this ad is not delivering". Same audience-size story as ABM. Static Image 1 is also PAUSED in the other five ad sets.
Static Image 3: $28.79 a lead on 2.8% of 28-day spend, against $45.11 and $30.24 for the two videos that take 76% of it. That is the whole reason the three concepts below are statics, and it is the reason to weight new budget toward the image ad names rather than adding more of them.
[01]-[10] are ACTIVE in 6 Lead Magnet ad sets and in ABM, and have taken $1.74 total. [09] Image Feed - Comp set score is the only one with a result: 1 lead from 23 impressions. That is noise, not a signal, and it will stay noise until the main campaign has budget to spend.
5. Competitor scan (Meta ad library, 23 Jul)
The Ads Library API rejects this app's token (code 10, needs identity verification), so this was pulled from the public library UI in a headless browser. What it shows:
- Triptease, Sojern and Cendyn are not running B2B ads to hoteliers at all. Every active US placement under their names is a guest-facing campaign they run on behalf of a client hotel (Alma Resort, The Views Madeira, Best Western De Anza, The Chedi Lustica, TradeWinds). They are buying demand for hotels, not selling to them, on this channel.
- Canary Technologies is the one real B2B advertiser in the set. Its pattern: direct address ("Dear hotels", "Hoteliers..."), a hard stat as the hook ("up to 40% of hotel calls go unanswered"), a named-property customer quote, an emoji checklist, and a trust number ("20,000+ hotels in 100+ countries"). Offer is always a demo.
- Duetto runs portfolio/enterprise language - "See demand before your competitors do", "Master complexity. Scale performance." Offer is "arrange a meeting".
- The gap: nobody in this set offers a concrete free artifact. Every competitor CTA is a demo or a meeting. Sail's free comp-set report is the differentiated offer on this channel, and the report's own contents (competitor ad previews, rate spread, review counts) are the material for hooks nobody else can copy.
Angle taken from Canary and applied below: lead with one hard, specific number, not a promise. All three concepts do that on the image itself.
6. Three new concepts
All three are new images, not new copy on an existing asset. Built on the white brand (Figtree 800, one PT Serif italic clause, blue pill CTA, no logo, no URL), rendered at 1080x1080. Each went through an independent design-QA and creative-critic pass, and the fixes below were applied before this page was written. Nothing has been uploaded. Reply on Telegram with a number to have it built as a PAUSED ad.
7. Recommended actions
All spend decisions are yours. These are recommendations, nothing here has been executed.
ABM is taking 63% of Sail spend at $136 a lead against $36 in the main campaign, on an audience of about 1,000 people already served 9 times over. This is the highest-value change available and it does not need any new creative. If ABM must keep running, cap it at a fraction of the main campaign rather than the majority of it.
Both lead-magnet campaigns read $10/day. At that level the main campaign cannot spend into 17 ad names, which is why the ten variants you switched on have $1.74 between them. Either raise the main campaign back toward its pre-11-July level, or accept that no creative test in the account will produce a readable result.
Video 1, Video 2, Static Image 1 and Static Image 3 in Sail Lead Magnet - eblast audience (A+A off) are all at Meta error 4469003. They are not spending and not going to. Pausing that ad set is cleaner than leaving four permanent error states in the account.
Static Image 3 is the cheapest lead in the account on 2.8% of the budget. [06], [09], [10] and any of the three concepts above are the natural place to send incremental budget - but only after point 2 is resolved.
Any concept you name gets built as a PAUSED ad in the best-fit Lead Magnet ad set, square plus a 9:16 story in one ad, named [NN] Image Feed - <hook> so it reports as one row. Nothing gets activated by me at any point.
Generated 23 July 2026 by Sail Ads Manager. Read-only run: 12 Sail campaigns, 216 ads and 3 campaign insight windows were read; the Meta ad library was read from the public UI. No campaign, ad set, ad, budget, bid or status was created or modified. Client-hotel campaigns in act_305550522185383 were excluded from every query.
One note on the brief spec: today's task text asked for the Sail wordmark top-left on the creative. It is deliberately absent - you had it removed from ad creative on 22 July and that instruction is still current, so it has not been put back. Say the word and it goes back on.