When a buyer asks an AI assistant which product to choose, does the answer name you? Most teams cannot say, because nothing in their analytics records it. Noma is built to answer that question every day, and to help you change the answer.
Noma asks ChatGPT, Gemini, Perplexity, Claude and Grok the questions your buyers ask, records how each answer describes your brand, shows the sources behind it, and turns the gaps into work your team can do.
One answer now decides the shortlist
We built Noma because buyers increasingly ask assistants for recommendations, and a single answer names only a few brands. Search rankings do not decide which ones.
A buyer who once scanned a page of links now types a question into an assistant and reads one reply. If you are named in it, you are in the running. If you are not, the buyer may never hear of you, and your search dashboard will not show that it happened.
Ranking well helps an engine find your pages, but it does not guarantee a mention. Engines draw on different sources and disagree with each other. Teams were checking answers by hand, one prompt at a time, with no history and no view of which pages the engine relied on. We wanted a repeatable daily measurement and a clear path from what it shows to what you do next. That practice is usually called answer engine optimization, or AEO.
Track: see whether AI answers name you
Noma asks every tracked prompt on each engine once a day and records the full answer, so every number links back to what the engine actually said.
Daily answers from five engines
Each prompt runs from a fresh session with no personalization, and you can re-run any prompt on demand. Self-serve plans include ChatGPT, Gemini, Perplexity and Claude. Grok is an add-on ($49 a month for access, then credits per answer) and is included on the free trial and Enterprise.
The metrics
- Visibility: the share of answers that name you.
- Share of voice: how often you are named compared with competitors, as explained in what AI share of voice is.
- Position: where you appear in the answer.
- Sentiment: a 0 to 100 score, measured only on answers that name your brand.
- Cited sources: the pages and domains each engine drew on.
The numbers are only as good as the prompts behind them, which is why we wrote a guide to which prompts to track.
Understand: find out why the answer looks the way it does
Noma shows which sources shaped an answer, where competitors appear and you do not, and how you are described when you are named.
Sources, Earned and Owned
Cited sources are split between third-party coverage (Earned) and your own pages (Owned), so you can see whether to work on your site or on how you are covered elsewhere.
Gap analysis
On Growth and above, gap analysis lists the prompts where competitors are named and you are not. These are usually the most useful place to start.
Sentiment
Read how the answers that name you describe you, and spot descriptions that are out of date or wrong.
Act: turn gaps into work
Noma turns what it finds into tasks: a site audit, content drafts built from gaps, and an Actions page that collects next steps.
A 12-point site audit
The audit checks structured data, FAQ and Q&A schema, llms.txt and AEO signals, heading hierarchy, author bylines, content freshness, internal link health, image alt text, word count, Core Web Vitals, mobile friendliness, and HTTPS and redirects.
Content drafts, published by you
Noma drafts content from the gaps it finds: 10 drafts a month on Starter and unlimited on Growth and above. Your team edits and publishes every draft. Nothing is published automatically.
What is Agent Analytics?
Agent Analytics uses your server logs to show how AI systems interact with your site. You connect logs by webhook, Vercel log drain or access-log upload.
It shows which AI crawlers read which pages, the visits arriving from AI assistants, and the pages that are crawled but never cited. That last list points to pages an engine has read without using anything on them.
What does Noma integrate with?
Noma connects to Google Search Console and offers an MCP server for Claude, CSV export, an API and white-label reports.
- Google Search Console: connect your search data.
- MCP server: query your Noma data from Claude.
- CSV export: take your data anywhere.
- API: on Scale, Studio, Agency and Enterprise.
- White-label reports: on Studio, Agency and Enterprise.
What Noma does not cover
Noma does not track Google AI Overviews, Google AI Mode, Microsoft Copilot, Meta AI or ChatGPT Shopping. It does not offer SSO or a SOC 2 report today.
Who is Noma for?
Noma is built for brands that want to know how AI describes them, agencies that manage this for clients, and teams who would rather have specialists run it.
- Brands: Starter at $79 a month ($65 on the annual rate) for 1 project and 50 prompts; Growth at $299 ($249) for 3 projects with 60 prompts each, adding gap analysis and unlimited drafts; Scale at $799 ($665) for 10 projects with 100 prompts each, adding API access and perception and attribute tracking.
- Agencies: Studio at $499 ($415) for 10 client workspaces and Agency at $1,199 ($999) for 25, each workspace with 50 prompts, white-label reports and dashboards.
- Enterprise: custom pricing, unlimited prompts and projects, all five engines including Grok, a dedicated success manager, custom onboarding and an SLA.
- Managed service: Noma strategists run the platform for you.
Prices are in USD. The annual rate gives 2 months free, is still billed monthly, and can be cancelled any time. Current prices are always on the pricing page.
How to get started
- Run a free brand audit, with no account, to see how AI answers describe you today.
- Start the 3-day free trial: 25 prompts on all five engines. It is free to start, and a card is taken.
- Choose a plan on the pricing page when the trial shows you what to track.
If you have questions first, the FAQ covers the common ones, and the AEO guide explains the method behind the product.