What we report, and when
A short report every month, in two halves: what we did, and what the measurements show. Every quarter, a full re-measurement and a decision about whether the plan should change. This page sets out exactly what’s in each.
Two halves, kept apart on purpose
What we did is Execution. What moved is Outcome. We report both and we never add them together.
This isn’t a formatting preference. If a report blends the work into the results, a busy month reads as a good month. We’ve measured cases where the business with the strongest technical work wasn’t the one AI assistants named, while a fairly ordinary one led on both counts. Anyone who tells you the two track each other hasn’t measured it.
Every month
Two lists, on the same page, never combined into one figure.
What we did — Execution
- The work log for the month, item by item
- What’s queued for next month
- Anything we recommended that you decided not to do, recorded as declined
What we measured — Outcome
- Google Search Console: impressions, clicks and the queries you’re appearing for
- Your position in the map results, for the fixed set of ten questions
- The AI assistant results for the fixed set of twenty questions, scored on named and cited separately
- Anything that changed enough to be worth pointing out, in either direction
Every quarter
Once a quarter we re-measure from scratch and decide whether the plan should change.
- A full re-score across all three angles, measured from scratch
- A fresh comparison with other businesses in your area, measured the same way on the same day
- A revision of the question sets, so they keep pace with what you’re trying to grow
- A decision: continue, change direction, change level or stop
The quarterly meeting is where we act on the measurements. That’s deliberate: month to month the AI numbers are noisy, and acting on noise produces work that looks responsive and achieves nothing.
The full schedule
Everything we measure, how it’s classified and when it arrives.
| What | Type | Monthly | Quarterly |
|---|---|---|---|
| Work log — what we did | Execution | Yes | Yes |
| Google Search Console — impressions, clicks, queries | Outcome | Yes | Yes |
| Position in the map results (10 questions) | Outcome | Yes | Yes |
| AI assistant answers (20 questions) | Outcome | Yes | Yes |
| Full scoring across all three angles | Execution | — | Yes |
| Comparison with other businesses in your area | Reference | — | Yes |
Why the AI numbers move when nothing has
An assistant doesn’t give the same answer twice. The same question can name a business one month and not the next, with nothing changed on the site.
We could hide this by reporting a smoothed figure. We’d rather you saw it, because the alternative is a number that looks like progress and isn’t. The fixed question set is how we keep the noise comparable instead of pretending it isn’t there.
What’s not in the report
Three things we leave out on purpose.
- No forecast. We don’t predict rankings, citations, traffic or enquiries.
- No claim of cause. Execution and Outcome sit side by side and stay there.
- No vanity metric we can’t connect to your business. If we can’t say why a number matters to you, it doesn’t go in.
Common questions
- Can we get the report more often than monthly?
- The measurements wouldn’t have changed enough to be meaningful. If something significant happens between reports, we tell you then rather than waiting.
- Who writes the report?
- We do, and the managing director is on every engagement. It’s a short document meant to be read, not a dashboard export.
- What happens if the numbers go down?
- We say so, and at the quarterly reset we work out whether it’s noise, a change in how the systems work, or something we did. All three happen.
- Do we get the raw data?
- The measurements come from your own Google properties and from our question sets. You keep access to the first, and the second is set out in the report.
Where to go from here
How the numbers are produced, and what the program does with them.