+144 net new followers in July, a 6.9% gain and steady all month. The two clearest step-ups tracked the July 6 empathy post and the July 17 anniversary and Meg Bear board announcements. Reach was less top-heavy than June, so growth was smoother.
Two AI-in-hiring takes (the recruiter-empathy post and the Workday and Eightfold lawsuits) drove about 83% of total reach. Everything else clustered under 2K.
Each list ranks by a different signal: impressions for raw reach, saves for "I want to come back to this" intent, and engagement rate for resonance on smaller-reach posts. This month two AI-and-hiring posts carried reach, while the podcast clips and event promos held a warm audience but stayed small.
Who got seen the most.
B2B's strongest signal. Readers bookmarking to revisit.
Smaller-reach posts that overperformed for their audience.
Image led volume (6 of 18), but text carried both breakout posts. Video (all podcast clips) and two link Articles rounded out the calendar.
Text's average is inflated by the two AI-in-hiring posts. Strip those two out and text sits near image. Video and link Articles trailed all month.
We read each of the 18 posts and classified them by the job the post was doing. Five recurring patterns emerged. The breakdown shows volume and performance per pattern so we can see which deserve more calendar share next month and which to rethink.
Hot takes have by far the highest reach ceiling this month. Educational sits second, carried by the lawsuits explainer. Podcast, event promo and product news all stay small.
Saves are the strongest intent signal. Educational owns this metric on the back of the lawsuits post; hot takes come second. Event and product news barely register.
Five content patterns observed across 18 published posts.
| Pattern | Posts | Total imp. | Avg imp./post | Total saves | Top performer |
|---|---|---|---|---|---|
| Hot takes / contrarian | 3 | 43,589 | 14,530 | 10 | Why recruiters turned to AI (41.2K imp) |
| Educational / thought leadership | 3 | 16,966 | 5,655 | 18 | Workday and Eightfold lawsuits (15.2K imp) |
| Podcast (Recruiters On The Rise) | 5 | 3,065 | 613 | 2 | ROTR: Stacy Donovan Zapar (1.2K imp) |
| Event & appearance promo | 6 | 2,508 | 418 | 1 | Seattle happy hour (595 imp) |
| Product & company updates | 1 | 1,498 | 1,498 | 0 | Welcoming Meg Bear (1.5K imp) |
Three contrarian posts drove 43.6K impressions, 65% of the month's reach. Each named real companies and took a stance: the recruiter-empathy post ("why isn't anyone asking what pushed them to it") hit 41.2K with 105 comments, and the Great AI Rehire post named Klarna and IKEA. This is the pattern Colleen can repeat reliably, and it is where the ceiling lives.
Three educational posts drove 18 of the month's 31 saves, more than any other pattern, on just 25% of the reach of hot takes. The Workday and Eightfold lawsuits post is the model: a named, current event, then "here are 4 questions I'd ask," which readers bookmark. Educational converts attention into intent better than the debate posts do.
Five Recruiters On The Rise video clips averaged 613 impressions and drove 2 saves. They keep a respectable engagement rate (3 to 4%) from a small following, but video underperformed on reach all month and the clips rarely get bookmarked. They occupied 28% of the calendar and returned 5% of the reach.
Six event and appearance posts (five Seattle happy hour variants plus the Hung Lee webinar teasers) averaged 418 impressions and one save combined. The audience saw the same event announced repeatedly. One or two strong invites per event would hold the reach; five dilutes it and trains the feed to scroll past.
The first sentence does almost all of the work on LinkedIn. The algorithm decides reach within the first 90 minutes based on early dwell and engagement, so the hook is where the post is won or lost. We analyzed the opening lines of Colleen's top 6 posts (top 5 by impressions and top 5 by saves, overlap counted once). Five ingredients recur.
Observed across the top performers this month.
"1,700 applications," "160,000 job applications in three months," "9.7% of AI-interviewed candidates got offers versus 8.7%," "half the companies that laid off workers are now hiring them back." Every breakout dropped a hard number before sentence three. Numbers create instant credibility and stop the scroll.
Workday, Eightfold, Klarna, IKEA, SAP SuccessFactors, University of Chicago. The winning hooks cite real names as evidence. Generic phrases like "companies today" or "the market" show up in the weaker posts.
"Recruiters are getting blamed for using AI to screen applicants. But why isn't anyone asking what pushed them to it?" "Contrary to popular belief, using AI in recruiting can actually help." Naming a common belief and then breaking it is the single most consistent trait of Colleen's high-reach posts.
"Here's what the data shows," "Here are 4 questions I'd ask," "Here's why." The strongest hooks tell the reader an explainer is coming, which buys the dwell time the algorithm rewards and gives readers a reason to save.
"My recruiter friend recently spent 3 whole weeks reading resumes," "My honest take is." Personal voice, fast opening line, no corporate warmup. The event and podcast posts that opened with "Final call" or "This new episode is live" never broke out of the warm audience.
Setting aside the two AI-in-hiring breakouts, the floor of the calendar was a boilerplate webinar post, a guest-intro podcast clip, and a repeat event teaser. Hypotheses below.
This post opened by repeating its own title twice ("Essential Guide for Increasing Talent Density. Essential Guide for Increasing Talent Density") then ran dense, jargon-heavy webinar marketing ("architecting environments where elite talent multiplies collective output"). It reads like generated copy, not Colleen. There is no number in the first line, no named company, no tension, and no first-person voice, which is the exact opposite of what worked this month. It also posted the same day as a podcast clip, splitting attention.
Opens with "Guillermina Cabello has been recruiting IT talent for more than 30 years." A warm line, but it leads with a person the audience does not know yet and uses video, which underperformed every week this month. Podcast clips need a thesis hook (the single most counterintuitive thing the guest said) in the first line, not a biography. It also shared the calendar with the talent density post above.
This was one of five Seattle happy hour posts inside a two-week span. By the fourth touch the feed had already seen the invite, so reach collapsed to a fraction of the first announcement. Event promo works on one strong invite and maybe one final-call reminder, not a weekly drumbeat of near-identical posts.
1. Boilerplate or generated-sounding copy with duplicated headers (the talent density post, worst of the month at 190).
2. Podcast clips that open with a guest biography instead of a standalone thesis.
3. Over-promoting one event across five near-identical posts (Seattle happy hour reach decayed with each repeat).
4. Reading too much into the 41.2K empathy post. It was driven by 105 debate comments and returned only 8 saves. Judge August reach against the ~1.5K non-viral median and lean on educational posts for intent.
Five concrete moves, each grounded in something we observed this month.
The AI-and-hiring hot takes are the reach engine. Aim for roughly 2 per week, each naming real companies and taking a clear stance (vendor lawsuits, layoff-then-rehire cycles, ghost jobs, AI screening). Space them at least 3 days apart so they do not compete with each other.
The Workday and Eightfold lawsuits post is the template that wins saves: take a current, named event and end with a numbered list the reader can act on ("4 questions I'd ask a vendor"). Educational drove 18 of 31 saves this month on a quarter of the hot-take reach. This is the intent engine.
The 5 ROTR clips averaged 613 impressions and 2 saves, most opening with a guest introduction. Reframe each around the single most counterintuitive thing the guest said, put that in the first line, and move the guest bio and episode link down. Test a text-plus-quote format instead of native video, which trailed all month.
The Seattle happy hour ran across five posts and the Hung Lee webinar across three, averaging 418 impressions with one save combined. Consolidate to one strong invite (with the hook and the human reason to attend) plus one final-call reminder. Redirect the freed slots to hot takes and educational posts.
The empathy post pulled 41.2K impressions but only 8 saves, because a debate invites comments, not bookmarks. The very next AI post that offered a checklist (the lawsuits explainer) doubled the saves on a third of the reach. When a debate post goes big, follow it inside 3 days with a framework the same reader would save.