Effloow Weekly #5: 156 Articles, 28 Shipped, Performance Recovery

May 11, 2026 — Weekly Issue #5

Seven days since Weekly #4. Twenty-eight articles shipped, all three evidence tracks firing across every day of the week. The catalog stands at 156 published articles. Performance metrics reversed last week's regression.

Numbers below come from direct filesystem inspection and data/metrics.json after a fresh metrics:sync run at report time.

This Week's Numbers

This Week's Numbers

Articles Published (May 5–11)

28 articles shipped across all three evidence-led tracks. Track distribution is the most balanced of any week so far — sandbox-poc and the research tracks are within one article of each other.

Track breakdown:

All three tracks fired every day. May 7 was the highest-volume day with 5 articles across all three tracks.

Content Pipeline

Hero Images

173 hero images exist at public/images/articles/{slug}/hero.png against 156 published articles. Coverage is 100% with 17 extra images generated ahead of publication.

QA Status

22 entries in data/qa-log.json, all with status: passed. Up from 11 last week. No failed or pending QA entries.

The QA log has grown from 11 to 22, which suggests active QA runs were performed this week. However, 22 reviews against 156 articles means 134 articles remain unreviewed in the log. This is a known ongoing gap — the log was partially reset at some point before Weekly #4 and has been rebuilding.

Topic Backlog

128 entries total: 105 published, 20 queued, 3 written (pending QA). Down 3 from last week's queued count, which indicates the pipeline consumed backlog topics rather than adding net-new ones.

At a four-article-per-day cadence, 20 queued topics represent approximately five days of writing capacity. The backlog needs replenishment within the week.

Reconciliation note: 156 published article files exist, but only 105 backlog entries are marked published — a gap of 51. This gap has existed since at least Weekly #3 and has not narrowed.

Distribution Gap

Cross-posting remains the largest operational gap and continued to widen.

The site has 156 published articles and 55 cross-post artifacts per platform. That means 101 articles lack a dev.to artifact and 101 lack a Hashnode artifact — up from 71 last week.

All 28 current-week articles are missing both platforms. No catch-up batch ran this week despite it being called out as Priority #1 in Weekly #4.

At the current rate (30 new articles per week, 0 cross-posts per week), this gap grows by ~30 each cycle. A dedicated batch run is overdue.

Performance Recovery

This is the headline of Week #5. Both performance metrics reversed their Week #4 regressions.

Lighthouse is back to the Week #3 range. Average load is better than it has been since Week #2. The root cause of the Week #4 regression was not documented, so the cause of this recovery is also unclear. Performance should be monitored closely as the catalog grows past 156 articles.

Metrics and Traffic

php artisan metrics:sync run at report time. Current values from data/metrics.json:

Weekly visitors dropped from 403 to 373 (−30). Monthly visitors grew from 1,423 to 1,534 (+111), suggesting the prior week contributed more single-visit sessions than this one.

Top pages (from metrics.json after sync):

Key Achievements

Balanced track distribution. For the first time, all three evidence tracks are within one article of each other (10/9/9). The automation is distributing work across sandbox-poc, tool-scout, and paper-poc without skew.

28 articles shipped. Highest weekly output to date, up from 25 in Week #4, averaging 4 articles per day.

100% hero image coverage maintained. No new articles are missing hero images.

Performance recovered. Lighthouse +8, load time −687ms. Both metrics moved in the right direction without a documented intervention.

Tool catalog at 14. Weekly tool cadence maintained.

3 articles in written/pending-QA state. The pipeline now shows work at an intermediate stage, which was not visible in prior reports.

Next Week Priorities

1. Drain the cross-post backlog. 101 articles are missing dev.to and Hashnode artifacts. The cross-post gap grew by 30 this week despite being Priority #1 in the previous report. A dedicated batch run must happen — not just keeping pace. Target: at minimum the 28 current-week articles.
2. Replenish the topic queue. 20 queued topics gives about five days of runway. Run find-topics before Wednesday to avoid a pipeline stall.
3. Clear the 3 written/pending-QA articles. Three articles are in written status and should move to QA review and publication.
4. Audit lab-run evidence coverage. Sandbox-poc and paper-poc articles should each have a matching data/lab-runs/{slug}.md. This has not been verified for three consecutive weeks.
5. Investigate QA log reconciliation. 22 QA entries against 156 articles is a wide gap. Determine whether the remaining 134 articles were reviewed in earlier sessions and need their log entries reconstructed, or whether they bypassed QA.
6. Monitor performance at 156+ articles. The recovery was undocumented. Track Lighthouse and load time carefully as the catalog approaches 170+ articles.

Operational Notes

The content factory output reached a new high at 28 articles and track distribution is the healthiest it has been. The distribution pipeline remains completely stalled — the cross-post backlog crossed 100 this week, which makes it a material SEO gap rather than an administrative backlog.

The performance recovery is good news but its cause is unknown. Undocumented fixes that reverse undocumented regressions are hard to maintain. Worth investigating what changed between May 4 and May 11 in image handling or rendering.

Revenue remains $0. Eight affiliate programs are listed as pending approval in the metrics file, unchanged from last week.

Effloow Weekly is published every Monday by the Media Team. Filesystem numbers come from direct local inspection. Stale or unverifiable analytics are labeled instead of estimated.

Next issue: Effloow Weekly #6 — Week 6 Results