GrowthBook LinkedIn Topic Review (organic · boosted excluded · all-time)

Ranked by median engagement per post. Confidence flags: <5 posts = low-confidence; <3 posts = insufficient (not actionable).

1. Top 3 best-performing topics by median engagement

Restricting to rows with usable samples (post_count ≥ 5):

1. Graham (CEO) — Industry commentary: median 18 across 8 posts (low-confidence, n=8 is borderline but acceptable).
2. Graham (CEO) — Product functionality: median 16 across 10 posts. Note: trending declining over 30d.
3. Ashley (me) — Thought leadership: median 16 across 14 posts (solid sample). Also Ashley — Other (median 16, n=5, low-confidence) ties.

Honorable mention: Graham — Thought leadership, median 15 across 60 posts — the single most reliable high performer by sample size.

2. Top 3 under-exploited opportunities (high median, low volume)

These show strong medians but thin volume — worth scaling if the signal holds:

1. Ashley (me) — Customer story: median 15 across 6 posts (low-confidence). Strong per-post return on a topic Ashley rarely posts.
2. GrowthBook — Other: median 16 across 5 posts (low-confidence). Outperforms the company's workhorse topics, but "Other" is a grab-bag — hard to replicate deliberately.
3. Alyssa — Product functionality: median 6 across 6 posts (low-confidence); modest but it's her best-sampled topic above Marketing event / Other, and flagged as newly active.

Insufficient to recommend (n<3), noted for awareness only: Graham — Customer story (median 30, n=2), Graham — Marketing event (median 30, n=2), Alyssa — Product demo (median 20, n=2), GrowthBook — Industry commentary (median 31, n=1).

3. Candidates to cut (high volume, low median)

• GrowthBook — Marketing event / content: median 8 across 50 posts, and declining. Highest-volume topic on the page with a low payoff — prime for trimming.
• GrowthBook — Product functionality: median 8 across 28 posts, declining. Second-highest company volume, flat-low return.
• GrowthBook — Podcast: median 7 across 14 posts. Underperforms Ashley's own podcast median (11) despite decent volume.

(Alyssa — Marketing event, median 3 across 12 posts, and Alyssa — Other, median 4 across 14, are also low-yield but lower stakes.)

4. Notable rising / declining (30d)

• Rising: Ashley — Podcast (median 11, 63 posts) — the only clearly rising well-sampled topic.
• Declining: GrowthBook — Marketing event (50 posts) and GrowthBook — Product functionality (28 posts); Graham — Product functionality (10 posts).
• New: Alyssa — Product functionality & Product demo; Graham — Past experience & Data science; GrowthBook — Customer story. Treat "new" as newly-posted, not a baseline.

5. Caveat on confounds

Topic performance is heavily tangled with who posts it. "Product functionality," for example, reads as a top performer for Graham (median 16) but a cut candidate for the GrowthBook page (median 8) — that's an author/audience effect, not a topic effect. Several high medians also sit on n=1–2 samples, so a single viral post distorts the ranking. Don't generalize a topic across accounts.

What changed since the last update

Since the prior snapshot on 2026-10-09, the headline mover is ads. Ad clicks: 378 in the last 7d vs 236 prior (+60%), with CTR essentially flat at 1.2% (vs 1.17%) — so the gain is reach/volume-driven, not efficiency. The two Omni | TOFU | EG | TA-JF-NA_UK campaigns (TL, 127 clicks and DA, 114 clicks) drove most of it, both at a strong ~4.5–4.7% CTR. Ad window ends 2026-10-09, so figures lag a few days. On organic, the clearest post signal is Ashley — Podcast rising: median 10 across 10 recent posts (vs 10 posts prior, median 6) — a genuine trend by the ≥2/≥2 rule, and the one movement that justifies a recommendation. The GrowthBook page's Marketing event / content is declining: median 5 across 11 recent posts (vs 6 posts prior, median 8), reinforcing the cut case above. Pipeline and ROAS held steady.

On recommendations vs. prior_recs: The podcast rise supports leaning into Ashley's podcast cadence — a slight update from "Double down on Thought leadership for Ashley." Most of last week's "double down" recs (Customer story for Graham, Product functionality/Product demo for Alyssa, Product demo for GrowthBook) rest on new or single-post signals this period and are not yet supported — hold, don't act. The Gr