Is DM Champ Named in AI Answers?

Mention rate per engine

Share of engine runs whose answer named the brand, from 10 run tests.

Imported baseline: measured before the brand was tracked in the product, then imported with the engine version of every run. It is drawn as its own segment and never joined to the product trend line.

Google AI Overviews are fetched 3 times per prompt rather than 10, because the overview is a search result page, not a chat answer.

Mention rate over time

One point per engine per measurement date. Imported and engine change periods are marked.

Imported points are the baseline. No engine change has been recorded so far. A period is marked as one when an engine changes its model version or the control prompts move by more than 10 points, and it is then not credited to any fix.

Every prompt, every engine

The consistency label from the latest 10 run test, with the runs that named the brand.

STRONG, MODERATE, WEAK and INVISIBLE are the 10 run labels explained on the methodology page. A 0 of 10 counts only when every run finished and the engine searched the web.

Where the brand is missing

The prompts no measured run named it for, and any answer that got it wrong.

• white label AI sales agent for agencies: not named on ChatGPT
• best white label chatbot for agencies: not named on ChatGPT, Claude, Gemini
• ManyChat alternative AI: not named on ChatGPT, Gemini
• Instagram comment to DM automation tool: not named on ChatGPT, Claude, Gemini
• AI sales agent agencies can resell to clients: not named on ChatGPT, Gemini
• white label AI sales agent for MARKETING agencies: not named on ChatGPT
• white label AI software I can resell to businesses: not named on ChatGPT, Claude
• AI tools you can resell under your own brand: not named on ChatGPT, Claude, Gemini
• best white label AI platform to start an agency 2026: not named on ChatGPT, Claude, Gemini
• best white label AI agents for agencies, voice vs chat: not named on ChatGPT, Claude, Gemini

The fix log

Every change we shipped for this brand, dated, with the link and commit where there is one.

No fix has been logged for this brand yet. Each fix gets a dated entry here the day it goes live, so the next 10 run test can be read against it.

How this was measured

The same engines, labels and rules every customer gets.

Day 0 is an audit from 15 September 2026, 10 runs per prompt per engine,, imported into the product. It used earlier engine versions than the product measures with today, so it is drawn as a separate imported segment and kept off the product trend line.

The baseline ran hours after a home page rewrite, before the engines had read it again. The first product measurement therefore reflects every change since 15 September together, and it is also confounded by the engine version change. We describe it exactly that way.

Two prompt sets are tracked side by side: one from June and one from September, each prompt with a stable key so the two can be compared over time.

Each engine is asked through its API with web search or grounding switched on, the way a buyer's question is answered today. A run counts as a mention only when the answer names the brand itself, not an echo of the question and not a different company with a similar name. The engine version shown is the model id the engine reported for that run.