Research · June 2026
We built a tool to measure AI brand accuracy. Then we scanned ourselves.
Before asking customers to trust Faro Index, we scanned our own brand across AI platforms and found classic identity confusion — models describing us as the wrong company — which is exactly the failure mode the product was built to detect.
The first question any honest buyer asks about a measurement product is simple: does it work on you? So before we asked anyone else to run a scan, we ran one on Faro Index. Here is exactly what AI said about us — and why it is the best demonstration of the problem we built the product to solve.
What the AI platforms said about us
When we scanned our own brand, the models did not describe an AI brand-monitoring tool. Claude described “Faro Index” as a data and analytics company focused on supplier and vendor risk management. ChatGPT treated “the Faro Index” as a financial metric. Neither answer was malicious or even unreasonable — they were the models doing exactly what models do: assembling the most probable answer from training data and public sources for an ambiguous, brand-new name.
This is identity confusion, and it is the single most common failure mode we see across the companies we scan. A name collides with an established entity — here, an unrelated firm and a finance-sounding term — and the model fills the gap with the wrong one. For a young company, this is not an edge case. It is the default.
Why this is a product insight, not a product failure
It would have been easy to quietly suppress our own results. We did the opposite, because the scan caught the exact problem the product exists to catch. Three things were true at once:
- Our name is ambiguous, so models defaulted to a more established entity and a financial term.
- We were new, so there was little authoritative, machine- readable content for the models to anchor to.
- Model knowledge lags. Even after we publish disambiguation content, Perplexity tends to update within days while ChatGPT can take weeks. Accuracy is a moving target, which is the whole reason it needs monitoring rather than a one-time check.
What we did about it
We treated our own brand the way we tell customers to treat theirs. We published explicit disambiguation content — who Faro Index is, what category we are in, and how we differ from the entities models were confusing us with. We tightened the schema and answer blocks on our site so models have machine-readable facts to cite. And we kept scanning, so we can watch the correction propagate platform by platform instead of guessing.
That is the loop the product is built around: measure how AI describes you, fix the on-site signals that cause drift, and re-scan to confirm the change landed. The point was never to be perfect on day one. The point is to know, precisely, what AI is saying — and to be able to move it.
Scan your own brand
See what ChatGPT, Perplexity, Gemini, and Claude actually say about your company today — and whether they are describing you, or someone else with a similar name.