Question 1

What does the AI Search Brand-Presence Diagnostic do?

Accepted Answer

It measures whether your brand shows up when buyers ask AI engines which vendor is best — the loss category that never appears in your analytics. You run a fixed prompt battery across ChatGPT, Perplexity, Gemini, Google AI Mode, and Claude, mark each engine 0/1/2 on presence, accuracy, and sentiment from what you actually see, and it returns PRESENT / PARTIAL / ABSENT per engine plus a rollup naming the weakest engine to fix first. It's a structured scorecard for your own observations — no scraping, no API, no monthly subscription.

Question 2

How is this different from the AI Brand Misinformation Watch?

Accepted Answer

They answer different questions in sequence. This measures presence first: are you in the answer at all, across the engines your buyers use? The AI Brand Misinformation Watch is about accuracy: when you do appear, is what the AI says about you correct, and how do you correct what's wrong? You can't have wrong facts about a brand that never comes up — so presence is upstream. This is the cheaper diagnostic you run first to find where you're invisible; the Watch is the deeper tool for fixing what's said once you're present.

Question 3

Why is an engine ABSENT even when I marked accuracy and sentiment high?

Accepted Answer

Because of the presence gate. If your brand doesn't appear in the answer (presence = 0), the engine is ABSENT regardless of the other two marks — you can't be described accurately or favorably in an answer you're not in, so those marks are moot. In the worked sample, Google AI Mode is marked absent but accurate and positive (a common reflex), and the gate correctly forces ABSENT, naming presence as the fix-first. Absence dominates: getting into the answer is the prerequisite for everything else.

Question 4

How is each engine's verdict decided?

Accepted Answer

By a presence-gated MIN. If presence is 0, the engine is ABSENT outright. Otherwise the verdict is the weakest of the three dimensions: all three at 2 is PRESENT, a weakest of 1 is PARTIAL, a 0 is ABSENT. So being present but described with wrong facts, or present but ranked below competitors, is PARTIAL — not PRESENT — because a flawed appearance still costs you the buyer. The rollup then names the engine where you're least visible as the one to work on first.

Question 5

Which engines should I track, and what if the list changes?

Accepted Answer

The kit ships with the five mainstream consumer surfaces in 2026 — ChatGPT, Perplexity, Gemini, Google AI Mode, and Claude — and the engine list is fully editable. Add Microsoft Copilot, Grok, Meta AI, or Google AI Overviews, or drop one your buyers don't use; track the engines that matter to your market. (Note that the Gemini app and Google AI Mode share a model but are queried as different surfaces, and AI Overviews is a third Google surface.) The method survives the next reshuffle because you control the list.

Question 6

AI answers change every time — is a one-time scorecard meaningful?

Accepted Answer

AI answers do vary run to run, which is exactly why the method is a fixed prompt battery you re-run periodically, not a single lucky query. Run the same prompts logged-out where you can, mark what you see, and re-run monthly or quarterly to track the trend — the direction (appearing more, or slipping out) is the signal, not any single answer. It's a deterministic scorecard for your own observations, and it guarantees no citation or ranking outcome. Not legal advice.

AI Search Brand-Presence Diagnostic

The loss with no analytics footprint.

Three dimensions, presence first.

Presence

Accuracy

Sentiment & framing

Score five engines and watch the presence gate fire.

A verdict per engine, and the one to fix first.

Anyone buyers might ask an AI about.

Measure presence, then fix it.

Straight answers before you buy.

Find out where the AI leaves you out.