Proof/Self-case
We are our own first case study, and we will say when it goes badly.
Momentence runs the same system on momentence.com that we sell. This page publishes the measurement framework now, and the numbers on the schedule below.
The short answer
This is a self-case, not client evidence. It shows that we operate our own system and that we are willing to publish the results, including the months where growth stalls. It is not a claim about what we achieved for someone else.
- Baseline set 2026-09-01, first published numbers 2026-11-01
- Seven metrics with fixed written definitions that do not change between months
- The keyword set and the 40-prompt set are frozen at baseline
- No numbers are published here until the real ones exist
What we measure
Seven metrics, one definition each.
Definitions are published before the numbers, so we cannot redefine a metric later to flatter a result.
| Metric | Definition |
|---|---|
| Organic sessions | Sessions from organic search in GA4, excluding branded queries where search console can separate them, reported monthly and rolling 90-day. |
| Indexed pages | Pages reported as indexed in Search Console, against total submitted in the sitemap. |
| Keyword coverage | Number of tracked commercial queries where momentence.com appears in the top 20, top 10, and top 3, on a fixed keyword set frozen at baseline. |
| AI citation share | Percentage of a fixed 40-prompt set, frozen at baseline, where Momentence is cited by name across the major answer and generative engines we track. |
| Form conversion rate | Unique form starts divided by unique page views, per lead-capture surface. |
| Qualified conversations | Booked calls that meet our published fit criteria. Not MQLs, not raw form fills, not newsletter signups. |
| Pipeline progression | Qualified conversations that advance to proposal and to signed engagement, with the source channel attached. |
Schedule
When numbers appear.
Baseline set
2026-09-01
Keyword set and prompt set frozen. Tracking validated before anything is measured.
First published numbers
2026-11-01
Two full months of movement against the baseline, published on this page.
Cadence after that
Monthly, published on the first business day of the month
Same seven metrics, same definitions, good months and bad months alike.
No numbers yet, on purpose.
The baseline was set on 2026-09-01. Publishing partial figures before two full months of movement would be exactly the kind of selective reporting this page exists to reject. On 2026-11-01 the numbers appear here whether they are flattering or not.
Experiment log
The AI visibility experiments running on our own site.
Each experiment states the question and the metric that answers it. Results publish with the monthly numbers.
| Experiment | Question | Metric that answers it |
|---|---|---|
| Answer-first restructuring | Do answer-first openers increase citation capture on commercial pages? | Citation share on the frozen prompt set, by page |
| Entity consistency cleanup | Does a single normalized category description raise category-prompt citations? | Category-prompt citation rate |
| Original benchmark research | Does first-party data earn third-party citations faster than opinion content? | Referring domains and citation mentions per asset |
| Published pricing | Does full pricing transparency raise or lower qualified conversation rate? | Qualified conversations per pricing page session |
| Free tool as an entry point | Does the technical hygiene snapshot convert better than a gated PDF? | Qualified conversations per tool completion |
| Ungating versus gating | Which produces more qualified conversations from the same content? | Qualified conversations per 1,000 sessions |
Commitments
The rules we are holding ourselves to.
- Numbers publish whether they are good or bad, including months where growth stalls.
- The keyword set and the prompt set are frozen at baseline so we cannot quietly swap in easier targets.
- Every metric definition on this page stays fixed. If a definition has to change, the change and the reason get published alongside it.
- This is a self-case. It is evidence that we run our own system, not evidence of what we did for a client.
The research methodology we hold our published studies to is on the AI Discovery Technical Hygiene Benchmark, and the reporting format we use with clients is in the sample monthly report.
Keep reading the operating system
Every artifact lives in the proof library. The engagement itself is documented in our methodology, the commercial terms are on pricing, and the fastest paid entry point is the Growth System Diagnostic.
Want the same reporting discipline on your program?
Start with the free AI discovery technical hygiene snapshot, or book 30 minutes and we will walk your first 90 days concretely.