Methodology · How the Visibility Score works
You should be able to check our work. Here is how to.
Every number you get from Forever Cited comes from a method that is written down, versioned, and run the same way for every firm: the questions your clients ask, five AI engines plus classic search, one 0–100 score, a baseline frozen on day one, the same questions re-asked every week. Nothing on this page is a promise about outcomes. All of it is a promise about measurement.
Composite method v1.0 · Re-checkable by anyone with the prompt set
One score, three parts, fixed weights.
Anything we haven’t measured yet is left out of the average, never counted as a zero — so a score can’t rise just because we connected a tool.
Step one
We start with the questions your clients actually ask.
Your prompt set is built for your specialty and your market — the way a real person types them, not the way a marketer would. “Who is the best employment lawyer near me for a wrongful termination case.” “Fee-only fiduciary advisor for a five-million-dollar portfolio in Denver.” “Spine surgeon in Scottsdale who does minimally invasive fusion.” Twenty of them for the baseline audit, expanded as the engagement runs, and kept under version control so a result from March can be compared with a result from September.
Every prompt is run on ChatGPT, Google AI (AI Overviews and AI Mode), Claude, Gemini and Perplexity. We record the full answer, not a summary: which firms were named, in what order, with what sources cited. You see those answers in your report. If your firm is absent, you see who was there instead.
Alongside the AI engines we measure classic search — rankings, brand-search ownership, indexability — and your local and entity record: Google Business Profile, directory profiles, reviews, name-address-phone consistency. AI engines draw on all of it, so all of it is in the score.
Step two
One 0–100 score, with the arithmetic on the table
The Visibility Score is a weighted average of three categories. The weights are fixed in code, versioned, and identical for every firm we measure — yours, your competitor’s, and our own.
AI visibility — 40%
The largest weight, because it is what you are buying: whether the five engines name your firm for your clients’ questions, the size of the citable footprint they can draw on, and the answer content you own. A firm can be technically perfect and still invisible here. This category is where that shows.
Search — 35%
Whether search engines can index you at all, whether your structured data tells them who you are, whether you own the results for your own name, where your target terms rank, and the strength of the sites linking to you. Answer engines lean on these signals; firms on page one are named far more often.
Local and entity — 25%
Your Google Business Profile, your presence in the local pack, your review volume, whether your name, address and phone match everywhere, and how many authoritative directories list you with verified data. This is the record an AI engine checks before it trusts you exist.
Three rules keep the number honest. A KPI we have not measured yet is excluded from its category average, never counted as zero — otherwise your score would jump the week we connected a tool, and that rise would be ours, not yours. A category with nothing measured is dropped and the remaining weights are renormalized, so the score stays on a 0–100 scale instead of quietly deflating. And the method is versioned: if the weighting ever changes, every active baseline is restated on the same pass and the previous value is kept in the record. Two firms are never graded on two curves.
Step three
Your baseline is frozen on day one. We never touch it.
Your first full measurement becomes the baseline. It is stored as an immutable record — the software will not let anyone, including us, edit it — and it is the fixed point every later report is compared against. If we ever restate a baseline because the method itself changed, the original value is kept beside the restated one and the reason is written down.
Why it matters to you: any agency can show you a number going up. A baseline you can point to, that we cannot change, is what makes the number mean something.
Step four
Every week, the same questions. Every answer, kept.
The prompt set is re-run weekly against all five engines and classic search. You get the actual answers side by side with your baseline: which engines name you now that didn’t, which prompts moved, which didn’t. If nothing moved, the report says nothing moved.
Why a band, not a point. AI engines answer the same question differently from one run to the next. A single run is a sample, not a fact, so we report your visibility as a range with the individual runs behind it. A firm that quotes you a single decimal from one run is quoting noise.
The rule that shapes everything else
One firm per specialty, per market.
For firms that compete nationally, the market is the segment they compete in.
When you sign, every competing firm in your specialty and market is turned away, in writing, for as long as you are a client. We define “specialty” and “market” with you in the agreement — a practice area and a metro for most firms; a named segment for firms whose competitors are national — so there is never an argument later about what was promised. It is the reason our measurement can be honest: we are never measuring you against a firm we also work for.
What we will not do to move your number
Built so hiring us is the low-risk move
!No guaranteed outcomes
Nobody can promise what an AI engine will say about your firm. Anyone who does is lying to you. We promise measurement you can re-check and work you can see.
✓Nothing publishes without you
Every page, article and profile is drafted for your approval and carries the disclosures your profession requires. Your advertising rules are part of the workflow, not a review at the end.
◎No fake signals
No bought reviews, no review gating, no automated link schemes on your domain, no statements attributed to you that you did not write. Signals that can be faked are signals that get discounted.
⊞No access to your client files
We measure and build from public information and what you choose to give us. Case files, patient records, client financials stay where they belong: with you.
⌂Nothing you would lose if we left
Every asset and every account is yours. Every automation we build lives in systems you own. Leave any time; it all stays.
◆No number we can’t explain
Ask how any figure in your report was produced and you get the prompt, the engine, the date and the answer — not a slide.
The free audit
See your score, built this way, for free.
Your 0–100 Visibility Score measured exactly as described above, the actual AI answers for your specialty and market, and your three biggest gaps — free, in a 30-minute readout. One audit per firm.