AI answers vary between runs, users and days, so no single check tells you much. What works is a fixed set of buyer questions, asked the same way on a schedule, with results recorded consistently. Over a few months the trend becomes clear even though each answer is noisy.
Short on time? Comppound does this work for SaaS companies. AI-native, and a person reviews every page.
Have us do itBuild a fixed question set
Write twenty to forty questions across the buyer stages from the foundations track. Keep the wording fixed. Include category questions, competitor questions and questions about specific jobs. Ask each in the same assistants every month.
Record the same fields every time
For each question and assistant, record:
- Whether you are named
- Who is named, in what order
- Which sources are linked
- Whether the description of your product is accurate
Comppound can do this for you
Want to be named in AI answers?
Comppound checks what ChatGPT, Perplexity, Gemini and Google's AI features say about your category, opens your site to their crawlers and builds the pages and mentions they cite. AI-native, with every page reviewed by a person.
- AI visibility baseline for your buyer questions
- Crawler access and quotable pages
- Monthly tracking of where you are named
Read AI referral traffic
Some AI products send visitors with a referrer you can see in analytics, such as chatgpt.com or perplexity.ai. Create a channel group or segment for them. Many AI-driven visits arrive with no referrer and look like direct traffic, so treat the count as a floor.
Turn findings into work
Each gap points to a lesson in this track. Not named and no source of yours linked: you probably lack the page. Named but described wrongly: fix the page and the third-party listings. Competitor named from a review site: work on that source.
Worked example · an illustration, not a client result
Tallyroom's monthly sheet
Tallyroom kept one row per question and assistant, and one column per month. Patterns showed up within a quarter: it was named for integration questions once those pages existed, and still missing from category questions.
| Question | Assistant | Named? | Sources linked |
|---|---|---|---|
| time tracking that syncs with xero | Perplexity | Yes | Tallyroom + Xero page |
| best time tracking for agencies | ChatGPT | No | Two review-site lists |
| timewell alternatives | Gemini | Yes | Tallyroom alternatives page |
Do it yourself
Start your baseline
- 1Write twenty buyer questions and freeze the wording.
- 2Ask each in three assistants and record the four fields.
- 3Create an AI referrals segment in your analytics.
- 4Set a monthly reminder to repeat both.
- 5Pick one gap and assign it to a lesson in this track.
Free tool for this: Free AI visibility report, AI Overview traffic loss calculator
Common mistakes
Checking once and reacting
A single answer is noise. Wait for a pattern across several runs.
Changing the questions
If the wording changes, the trend means nothing. Add new questions alongside, do not edit old ones.
Counting only clicks
Many AI answers name a product without sending a click. Being named matters even when analytics shows nothing.
Rather not do it yourself?
Want to be named in AI answers?
Comppound checks what ChatGPT, Perplexity, Gemini and Google's AI features say about your category, opens your site to their crawlers and builds the pages and mentions they cite. AI-native, with every page reviewed by a person.
- AI visibility baseline for your buyer questions
- Crawler access and quotable pages
- Monthly tracking of where you are named
Questions
Yes, several track prompts across assistants. A spreadsheet works fine to start, and teaches you what to look for before you pay for one.
Monthly is enough for most SaaS companies. Answers change, but not so fast that weekly checks add much.
Key terms
- AI VisibilityAI visibility is how often and how accurately your brand shows up in answers from AI tools like ChatGPT, Perplexity, Claude, Gemini and Google AI Overviews.
- AI Share of VoiceAI share of voice is how often your brand appears in AI answers for a set of prompts, compared with how often your competitors appear.
- Prompt TrackingPrompt tracking is regularly running a fixed set of buyer prompts through AI assistants and recording whether, where and how your brand appears.
- AI Referral TrafficAI referral traffic is visits that arrive from AI assistants and answer engines, such as ChatGPT, Perplexity, Gemini, Copilot and Claude, when users click a cited link.
- AI Answer VariabilityAI answer variability is the way the same prompt can produce different answers, sources and brand mentions across runs, users, places and time.
- Dark trafficDark traffic is visits from sources analytics cannot see, such as private messages, apps and documents, which usually show up as Direct.
- Channel groupingChannel grouping sorts traffic sources into buckets like Organic Search, Paid Search, Email and Referral using rules based on source, medium and campaign.
