The weekly SEO audit

Once a week the system benchmarks every tracked keyword against its baseline and its position seven days ago, scores the movement, and returns a one-word verdict: baseline, improving, flat or declining.

What it is

A stored report with a full metrics snapshot — authority and stage, rank buckets, a per-keyword benchmark of baseline → last week → now → target, traffic this week versus last, posts published against the cadence target, and open site-health errors. Movement is scored from weighted signals (top-10 count, referring domains, keywords entering the top 100, climbs versus drops, session change) and the sign of that score is the verdict. The very first run is stored as the baseline every later report measures against.

Also called: SEO report · weekly benchmark · am I improving · SEO verdict

See it
The weekly SEO audit
Won
37%
Avg cycle
11d
Pipeline
$389k
JanSep
A report card: the verdict word, the signals list, and the benchmark table with baseline / now / 7-day arrow / 'to top 10' per keyword. Rendered by src/app/platform/seo/page.tsx. Sample data — no customer information appears here.
How it works
  1. 1For each tracked keyword the earliest recorded position becomes the baseline, the last reading before seven days ago becomes the prior, and the newest is current.
  2. 2Buckets classify every keyword into top 3, top 10, striking, page 2+, deep or unranked.
  3. 3Traffic is rolled from web_events for the last seven days and the seven before it.
  4. 4Signals are weighted — top-10 change ×3, ranked-count change ×2, referring domains capped at ±3, climbs minus drops, session change — and summed into a score.
  5. 5verdict = baseline when there is no prior report, otherwise improving / flat / declining by the sign of the score.
  6. 6The report is written to seo_reports with its proposals and a status of pending, or implemented when there is nothing to propose.
Why we built it

The module header: "The SEO audit loop (Keith 08-20): baseline once, then a weekly benchmark — what pages rank for what, how far from target, improving or not. When we're NOT moving, the report carries corrective proposals grounded in the data." The point of the baseline is that a marketing programme without one can always be described as working. Recording the starting position once, then measuring against it, makes 'is this working' a question with an answer.

The problem
  • No baseline existed, so improvement could always be claimed.
  • Reports described activity rather than movement.
  • A flat or declining week produced no specific corrective action.
Sound familiar?
What you get
One word — improving, flat or declining — backed by named signals
Every tracked keyword shown as baseline → now → distance to top 10
A bad week produces specific proposals instead of a discussion
What's inside

See it on your own jobs

Twenty minutes, your numbers, no slide deck. We’ll build one of your real buildings in front of you and send you the estimate link at the end — yours to keep either way.

or keep browsing features →