| Date | Hook | Impr. | Eng. | Eng. rate | C:L | Read |
|---|---|---|---|---|---|---|
| Jul 13 | Square just put restaurants inside ChatGPT and Claude, with no marketplace commission. | 17,940 | 52 | 0.29% | 1.08 | News + falsifiable claim + operator anchor. The model post. |
| Jul 28 | Monday.com cut 20 percent of its staff and told everyone the story was AI. The real story is on the price tag. DISTRIBUTION TRAP | 23,513 | 22 | 0.09% | 0.22 | News-reactive, huge reach, 69 link clicks - but rate cratered. |
| Jul 10 | I have never won a fractional client from outbound. Not one. | 598 | 13 | 2.17% | 0.30 | Fractional-entry pattern + hard claim. Best rate of the week. |
| Jul 30 | I ran a 50 million dollar media budget across five dating brands. The company still went insolvent. | 467 | 7 | 1.50% | 0.00 | Falsifiable retention claim; solid rate on modest reach. |
| Jul 23 | A founder once slid my six-pager across the table and said, that is pure fantasy. | 1,042 | 5 | 0.48% | 0.00 | Founder-trust narrative; reach without a hook to argue with. |
| Aug 6 | The Scrum Guide never uses the words product manager. | 314 | 4 | 1.27% | 0.00 | Sharp POA/PM test, but no news hook - low ceiling. |
| Jul 20 | Expedia just signed Allegiant to a 12-month exclusive. That was the last holdout. | 1,399 | 4 | 0.29% | 0.00 | News-reactive marketplace; distributed but thin engagement. |
| Aug 4 | I keep hearing that the agents a team builds are its moat. Mine have a shelf life of six to nine weeks. | 319 | 2 | 0.63% | 0.00 | Strong claim, thin distribution this cycle. |
| Aug 3 | Amazon is about to require sellers to label AI-generated people in product photos. | 379 | 2 | 0.53% | 1.00 | News-reactive, provenance angle; drew 1 comment. |
| Jul 27 | The whiteboard had two words on it. Total. Qualified. | 259 | 2 | 0.77% | 0.00 | North-star metric story; no claim up top, muted. |
| Jul 21 | "I get in trouble if the Friday status report is late. Nothing happens if I never talk to a customer." | 494 | 2 | 0.40% | 0.00 | Org-diagnosis story; reactions only, no debate. |
| Jul 16 | The most effective fix for a stubborn behavior problem was moving who absorbed the cost. | 291 | 2 | 0.69% | 0.00 | Incentive-design insight; small reach, no claim in line one. |
| Jul 14 | At Acquia I watched an enterprise company try to flip to product led growth like a switch. | 501 | 1 | 0.20% | 0.00 | Longest post, no hook up top - lowest rate of the week. |
Square, no marketplace commission (Jul 13). 52 engagements on 17,940 impressions, and 26 comments against 24 reactions - the comment-to-like near-parity is the resonance tell. News-reactive hook (Square + ChatGPT), a falsifiable claim (the aggregator tax collapses), an operator anchor (EverQuote), and a direct positive close. This is the template.
Never won a fractional client from outbound (Jul 10). Best engagement rate of the week at 2.17% (13 engagements / 598 impressions), 3 comments. The fractional-entry pattern plus a hard, falsifiable claim in line one - "Not one." Highest resonance per impression even without algorithmic reach.
Monday.com pricing (Jul 28) - distribution trap. The algorithm pushed it hardest of any post (23,513 impressions) but engagement rate fell to 0.09%, below the 0.15% trap line. It did drive 69 clicks to monday.com and 4 comments, so it was not inert - but reach that far ahead of resonance signals a hook that traveled wider than it landed.
Acquia PLG graft (Jul 14). Lowest engagement rate of the week (0.20%, 1 reaction on 501 impressions). Longest post in the set, no falsifiable claim up top, a single dense narrative. When it is a story, the claim still has to lead.
The window's top performers match the stored profile on every confirmed pattern: news-reactive format dominant (Square, Monday.com, Amazon, Expedia), operator specificity with named companies and numbers, falsifiable claims driving comments, zero emojis, no hashtags, and direct positive closings with no "That is not X, that is Y" negation. No new pattern emerged across multiple top posts and no stored pattern disappeared from them - one strong window is not a shift. The stored profile (50-post refresh, Dec 2025 - Apr 2026) remains authoritative.
The same posts ran on X, Bluesky and Mastodon. Standalone engagement was negligible, so they are summarized rather than tabled:
| Platform | Posts in window | Total engagement | Impressions / eng. rate |
|---|---|---|---|
| X | 12 | 0 | Views public (4-35/post); rate effectively 0% |
| Bluesky | 31 | 11 | n/a - impressions not exposed |
| Mastodon | 21 | 5 | n/a - impressions not exposed |