Site & Blog
As of Sep 18, 2026, 4:14 AM ET
Six page-one URLs drew 382 search impressions in the last 28 days and returned zero clicks between them. That is the whole story this week: rankings keep improving, impressions are holding 44% above baseline, and sessions have fallen back to their summer floor. The lever is the snippet, not the content.
Priorities this week
2. Rewrite the intro on the CPO-role post. High /blog/the-chief-product-officer-role-changes-every-time-the-company-doubles/ is the largest blog arrival pool on the site at 53 sessions, and holds people 16 seconds. Clarity has it at 8 to 19% scroll across every early-September sample. The traffic is arriving and leaving before the argument starts.
3. Chase the script error. Medium Clarity has logged one in each of the last three sample windows, and Sep 14-16 shares no day with Sep 11-13, so at least two separate errors fired. The watch item read zero in every sample before Sep 10.
4. Internal-link the deep-read posts into the pages fixed in step 1. Medium /blog/openclaw-setup-guide/ (264s average), /blog/ai-tools-for-product-managers/ (99s), /blog/product-review-meeting/ (88s) and /blog/fractional-cpo/ (124s) hold attention and are the right places to send readers onward from.
5. Resume weekly publishing. Maintain The August lift to 438 sessions came from distribution, not search, and nothing has replaced it. AI-flagged content holds at zero across six consecutive crawls, so cadence costs nothing on that front.
Key numbers
+44% vs ~1,100 baseline
-42% week over week
Sep 14-16, 1 script error
clear across 6 crawls
270 URLs, crawl Sep 16
Search impressions, weekly
Seven of the last eight full weeks sit in a 1,535 to 1,644 band. The trailing point is a two-day partial week and is greyed.
Site sessions, weekly
Same weeks as the chart above. The August spike to 438 and its unwind to 127 both happened while impressions held flat, so neither came from search.
Page-level signal, last 28 days
| URL | Sessions | Avg duration | Search impressions | Signal |
|---|---|---|---|---|
| / | 262 | 24s | 159 | Top arrival, best converter at 6.92% CTR |
| /blog/the-chief-product-officer-role-changes-every-time-the-company-doubles/ | 53 | 16s | - | Leaking - fix intro |
| /blog/ai-agents-replacing-comparison-marketplaces/ | 49 | 76s | 69 | Holds attention |
| /blog/product-review-meeting/ | 40 | 88s | 84 | Holds attention |
| /blog/ai-tools-for-product-managers/ | 29 | 99s | 25 | Holds attention |
| /blog/directly-responsible-individuals/ | 21 | 20s | - | Skims |
| /blog/openclaw-setup-guide/ | 21 | 264s | 362 | Top engagement, 1.10% CTR |
| /media-kit/ | 20 | 14s | 82 | Skims |
| /blog/fractional-cpo/ | 17 | 124s | 527 | Holds attention, ranks 13.4 |
| /blog/career-ladder-examples/ | 15 | 6s | 167 | Leaking - fix intro |
| /blog/ | 9 | 1s | 38 | Index page, no fix needed |
Sessions and duration from GA4, impressions from Search Console over Aug 19 to Sep 15. A dash means the page drew no search impressions in the window, so its traffic came from elsewhere.
Search pages with volume and no clicks
| URL | Impressions | Position | What it needs |
|---|---|---|---|
| /blog/coaching-is-not-therapy-a-ted-talk-or-a-pep-talk/ | 190 | 4.3 | Title and meta rewrite - largest page-one pool returning nothing |
| /blog/paid-acquisition-trap/ | 65 | 5.5 | Title and meta rewrite |
| /blog/north-star-metric/ | 46 | 1.3 | Title and meta rewrite - ranks first and still converts nothing |
| /blog/ | 38 | 5.7 | Index page - counted in the 382 but not a rewrite target |
| /blog/sample-software-release-plan/ | 23 | 8.4 | Title and meta rewrite |
| /blog/embarcs-journey-from-chaos-to-collaboration/ | 20 | 8.9 | Title and meta rewrite |
| /blog/product-strategy/ | 1,787 | 55.4 | Internal links and depth - too deep for a snippet to matter |
| /blog/agile-maturity-assessment/ | 733 | 45.6 | Internal links and depth |
| /blog/software-development-team-structure/ | 335 | 54.6 | Internal links and depth |
The top six are the page-one zero-click set and sum to the 382 impressions in the lede. The ranking is already won on each and only the snippet is failing. The bottom three carry far more volume but sit past position 45, where a rewritten title changes nothing.
User behavior detail
21 stored 3-day samples, summing to roughly 1,102 human sessions. Cumulative friction across every sample that records it: 43 dead clicks, 10 quickbacks, 3 script errors, and zero rage clicks, error clicks or excessive scroll. Consecutive windows overlap by a day or two, so the session total is a running sum, not distinct visitors.
Sample history
| Window | Human | Bots | Dead | Quickback | Script |
|---|---|---|---|---|---|
| Aug 5-7 | 37 | 17 | not recorded | ||
| Aug 6-8 | 34 | 9 | not recorded | ||
| Aug 10-12 | 56 | 135 | not recorded | ||
| Aug 13-15 | 50 | 122 | not recorded | ||
| Aug 15-17 | 44 | 26 | not recorded | ||
| Aug 19-21 | 135 | 16 | not recorded | ||
| Aug 20-22 | 76 | 20 | not recorded | ||
| Aug 23-25 | 41 | 9 | 1 | 0 | 0 |
| Aug 26-28 | 58 | 12 | 16 | 1 | 0 |
| Aug 29-31 | 94 | 23 | 5 | 3 | 0 |
| Sep 1-3 | 59 | 18 | 1 | 2 | 0 |
| Sep 2-4 | 41 | 11 | 1 | 2 | 0 |
| Sep 4-6 | 39 | 13 | 0 | 0 | 0 |
| Sep 5-7 | 28 | 9 | 0 | 0 | 0 |
| Sep 6-8 | 21 | 6 | 0 | 0 | 0 |
| Sep 7-9 | 35 | 10 | 4 | 1 | 0 |
| Sep 8-10 | 44 | 13 | 5 | 1 | 0 |
| Sep 9-11 | 50 | 18 | 1 | 0 | 0 |
| Sep 10-12 | 46 | 15 | 3 | 0 | 1 |
| Sep 11-13 | 43 | 29 | 5 | 0 | 1 |
| Sep 14-16 | 71 | 34 | 1 | 0 | 1 |
Friction counts were not captured before the Aug 23-25 sample. Rage clicks, error clicks and excessive scroll read zero in every sample that records them and are omitted.
Page engagement, latest sample (Sep 14-16)
| URL | Sessions | Scroll | Active | Total open | Read |
|---|---|---|---|---|---|
| / | 16 | 44.7% | 27s | 45s | Moderate |
| /blog/paid-acquisition-trap/ | 7 | 97.0% | 89s | 670s | Deep read, then left open |
| /media-kit/ | 3 | 67.3% | 17s | 66s | Skim |
| /blog/how-to-improve-team-culture/ | 3 | 70.3% | 211s | 312s | Deep read |
| /blog/openclaw-setup-guide/ | 3 | 16.3% | 3s | 61s | Idle - left open |
| /blog/ai-agents-replacing-comparison-marketplaces/ | 2 | 39.5% | 82s | 127s | Moderate |
| /blog/fractional-model-origin-story/ | 1 | 100% | 26s | 26s | Moderate |
| /blog/career-ladder-examples/ | 1 | 29.0% | 27s | 27s | Skim |
| /blog/growing-ecommerce-revenue-0-to-200mm-at-tripadvisor/ | 1 | 23.0% | 0s | 0s | Skim |
| /blog/product-strategy/ | 1 | 14.0% | 1s | 1s | Skim |
| /p/a45cf04a-d473-4115-84e6-ffa9b7249055/ | 1 | 9.0% | 0s | 0s | Skim |
Per-session averages. Samples from Sep 3 forward use the current per-session fields; earlier samples use the legacy ones, so per-page values across that boundary were not measured identically. The openclaw guide is the page to watch: it holds the site's deepest engagement over 28 days and its deepest search pool, yet reads idle in this window.
Crawl Sep 16, 2026, health score 100 over 270 URLs. Search Console data through Sep 15, GA4 through Sep 17, Clarity sample window Sep 14-16. Weeks are Monday-start and aggregated from daily rows; trailing partial weeks are greyed and are not a level. The site-audit issue feed returned 100 rows and stopped part-way through the Links category, so warning-level counts reconcile at 78 of the 174 headline and notice-level at 43 of 109 - the actionable categories that returned all read zero, but the full issue list was not seen and the one error URL was not identified. Ahrefs audits in-body images only, so feature-image alt text is outside this crawl.