Visibility Diagnostics
Watch your visibility every month, before a competitor takes the answer
For med-spa, home-services, dental, and solo-legal owners who want their standing across search and AI answers measured on a cadence, not guessed at once a year.
Every engagement is directed by a technical specialist and reviewed before delivery.
What this is
Visibility Monitor is a standing monthly subscription that re-measures your Machine-Readiness Score and tracks your standing across four pillars over time: Classic Search, AI Answers and Share-of-Answer, Reputation and Sentiment, and Technical Foundation. Every cycle we re-run a frozen panel of your real buyer questions across the answer engines, re-read rankings and reviews, re-check technical health, and compare the new reading to the last one. When a pillar moves in a way that matters, whether you gained ground or lost it, an alert follows with the engine, the locale, and the date attached, plus a short plain-language note on what changed. It runs monthly because AI citation is not stable. Cited sources drift heavily month to month, so a one-time audit describes a position that may no longer hold. You end up with a dated, trend-line record of where you stand and early warning when the ground shifts.
The problem
Why this matters now
Most owners find out they have gone invisible in AI answers the same way a burst pipe gets discovered, after the damage is done. A competitor starts getting named by ChatGPT, a review dip quietly drops you out of the local pack, or a site change breaks your indexation, and nobody notices for months because nobody is looking.
The trap is that a single audit, however good, is a photograph of a moving thing. AI answers are not fixed, and the set of sources an engine cites for the same question changes constantly. Independent large-sample studies report that 40 to 60 percent of cited domains change month to month, and the majority turn over within six months (Yext and SISTRIX, 2026, as cited in the Raveneye share-of-answer article). A score you earned in March can quietly decay by June.
The real problem is not knowing the number once. It is holding a trustworthy read of the number over time, catching drift early, and being able to prove whether the work you are paying for is actually moving the surface or standing still.
How it works
The mechanism, made checkable
- 01
We freeze the panel once, then hold it steady
At setup, a specialist builds a fixed panel of your real buyer questions and the competitor set to watch against. We freeze the panel so that every month measures the same thing, which is the only way a month-to-month comparison holds up. This is the same frozen-prompt approach the Machine-Readiness Score is built on.
- 02
We re-read all four pillars on the agreed cadence
Each cycle we re-measure Classic Search presence, AI Answers and Share-of-Answer, Reputation and Sentiment, and Technical Foundation. The AI-answer pillar runs each buyer question across each engine many times, because a single check is unreliable, and reports your appearance rate as a range with a confidence band rather than a single confident number.
- 03
We stamp every reading with engine, locale, and date
All three change the result, so all three ride with every observation. A reading without them is not a measurement, it is an anecdote. This is also what keeps your trend line comparable from one month to the next.
- 04
We compare the new reading to the last cycle and compute the delta
We set the new reading against the prior one, pillar by pillar and question by question, so movement is visible as a delta on a dated line, not a vibe. You see which surface gained and which slipped, and by how much, with the uncertainty band shown plainly.
- 05
Drift that matters triggers an alert
When a pillar moves beyond normal run-to-run variance, an alert follows: a new competitor appearing in the AI answers for your questions, a review-rating slide, a ranking drop, or a technical regression like a Core Web Vitals or indexation break. The alert names what changed and why it matters, in plain language.
- 06
We review it, then deliver it
Before anything reaches you, a technical specialist reviews the cycle to separate real movement from noise, because engines are volatile and a raw alert on ordinary churn would waste your attention. You receive the reviewed read, not a raw feed of alerts.
What is included
What is delivered
- A monthly (or weekly, at higher tiers) re-read of your full Machine-Readiness Score across all four pillars, human-reviewed before delivery.
- A frozen buyer-question panel, run repeatedly across the answer engines for a share-of-answer read reported as a rate with a confidence band.
- Classic-search presence tracking across your priority query panel and the local map pack, on a stated date and locale.
- Reputation and sentiment tracking across the review platforms that decide who gets chosen once found.
- Technical Foundation checks re-run each cycle: indexation, Core Web Vitals, rendering path, and schema health against published standards.
- Drift alerts when a pillar moves beyond normal variance, each naming the change and what it means.
- A named competitor set scored beside you every cycle, so movement is always relative, not just absolute.
- A dated cycle report with pillar deltas and a short plain-language note on what changed and what to watch.
- Every reading taken on a consistent, published method, with the comparison group disclosed.
The outcome
What it moves
- A dated trend line of your Machine-Readiness Score and its four pillars, showing the direction of movement, not just a single figure.
- Early warning when a competitor starts winning the AI answers for your buyer questions, while there is still time to respond.
- Proof of whether visibility work, from any provider, is actually moving your surface or standing still, measured the same way every cycle.
- A standing share-of-answer read across the engines that matter, reported with variance and stamped with engine, locale, and date.
- Technical regressions caught on cadence, such as an indexation break or a Core Web Vitals slip, before they quietly erode your presence.
- A running record that feeds cleanly into any build or retainer, so scoping the next fix starts from measured history, not a fresh guess.
What you get
What you get, and how it is priced
Visibility Monitor runs the exact measurement discipline documented in the methodology, on a repeating cadence, so your Machine-Readiness Score becomes a trend line instead of a one-off snapshot. Here is what runs in each cycle.
| Starter. The essential standing read for a single-location business. A monthly Machine-Readiness Score across all four pillars, a core buyer-question panel run across the primary answer engines, reputation and technical checks, and drift alerts. Same method as every tier, sized for one location and one core competitor set. | Quoted |
| Growth. For a business actively working its visibility. Everything in Starter with a larger frozen prompt panel, a wider engine set for the share-of-answer read, a broader named competitor set scored beside you, and more frequent re-reads so an active optimization push is measured closely rather than once a month. | Quoted |
| Scale. For multi-location or multi-service operators, or any business that needs the closest watch. The widest prompt panels and engine coverage, weekly cadence available, the fullest competitor and locale tracking, and the deepest specialist review of each cycle. Capacity and coverage differ; the measurement discipline is identical. | Quoted |
You see the full deliverables and cadence first, then a price built for your business, confirmed in writing.
Straight answers
Questions about Visibility Monitor
Is this run by an offshore team or produced automatically?
A technical specialist directs every cycle and reviews it before delivery. Technology handles the repetitive measurement so a person can spend their judgment on what actually changed, separating real movement from ordinary engine churn. We bill in USD, measured against the US engines and surfaces your buyers actually use. What arrives is a reviewed read, not an automated dump, held to the standard documented in the methodology.
Why does monitoring even matter? I already had an audit.
Because an audit freezes one moment of something that keeps moving. AI answers are not stable: independent large-sample studies report 40 to 60 percent of cited domains changing month to month, and the majority turning over within six months (Yext and SISTRIX, 2026, cited in the Raveneye share-of-answer article). A score you earned in one month can decay in the next without anyone touching your site. Monitoring is how a one-time position becomes a held one, and how a slip is caught while there is still time to act.
How is your measurement verified?
The method is published. The methodology page documents the four pillars, how we normalize raw signals to one 0 to 100 scale, and how the pillars combine, the same discipline recognized composite indices are held to. A technical specialist with years of hands-on work in search and AI visibility directs every cycle and reviews it before delivery. Every reading is stamped with the engine, the locale, and the date, and reported with variance, so you can check the working.
How is share of answer actually measured each cycle?
We freeze a panel of your real buyer questions and run each across each engine many times, because a single check is unreliable when answers vary run to run. We report how often you appear as a rate with a confidence band, and stamp every reading with the engine, the locale, and the date. It is directional sampling across a defined panel, not a census, because no engine publishes query volume or impressions, and we state that limitation plainly.
Why is it priced from $49 to $199 rather than scoped like your services?
Because it is a standard product with a fixed scope, so the price is published and work can start directly. The three tiers, Starter, Growth, and Scale, run the identical measurement method and differ only by capacity: panel size, engine and competitor coverage, cadence, and depth of review. Nothing here is priced for more rigor, only for more surface watched more often.
What is guaranteed?
Nothing about a ranking, a citation, or a traffic number. AI Overview selection is undocumented and volatile, engine behavior changes, and the click impact of AI answers is genuinely contested. We commit to repeatable measurement: the same frozen panel, the same method, reported with variance and dated, every cycle.
Does an llms.txt file boost the number you track?
No. An llms.txt file is AI-crawler readiness hygiene, not a proven ranking or citation lever. Google has confirmed its Search systems do not use it. Visibility Monitor measures your real presence in real answers, and it will not credit a self-declared file that engines do not act on.
Do I need an audit first, or can I start monitoring straight away?
Starting straight away is possible. Most owners begin with the free Machine-Readiness Score and the Surface Intelligence Audit so the monitor is pointed at a real, understood starting position, but that is a recommendation, not a gate. If you already know where you stand, you can start the trend line today.
Can I cancel, and do I own the readings?
Yes to both. There is no lock-in; the subscription runs month to month and cancelling takes the same number of steps as starting. The dated readings we deliver belong to you, and you can scope any future build or retainer from measured history rather than a fresh guess.
Related
Where this connects
Surface Intelligence Audit
Start with a full four-pillar teardown and a ranked fix list, then let the monitor hold the line the audit sets.
ExploreThe Machine-Readiness Score
See exactly how the 0 to 100 number and its four pillars are built, the same reading the monitor re-runs on cadence.
ExplorePricing
Where the monitor sits on the ladder, from a free Machine-Readiness Score to standing programs, with every product price stated plainly.
ExploreProvenance
Sources
- Aggarwal and colleagues, GEO: Generative Engine Optimization, KDD 2024, arXiv:2311.09735 (peer-reviewed, tier 1)
- Yext and SISTRIX, large-sample AI citation drift studies, 2026 (tier 3), cited in Raveneye Global, Share of Answer article
- Google Search Central, AI features and your website, on llms.txt and AI Overviews (tier 2)
Begin with where the business stands.
No obligation. The deliverable is a measured starting position and the corrections that move it most.