How to Track Your Brand in ChatGPT, Gemini, Perplexity and Claude

You can't improve what you don't measure, and AI answers are especially easy to misjudge from a handful of manual tests. Here is a repeatable method for tracking your brand across the major assistants, whether you do it by hand or with a tool like Citewise AI.

Step 1: Decide what "visible" means for you

Write down the outcomes you care about. For most brands they are:

• being mentioned when buyers ask category questions;
• being named ahead of specific competitors;
• having your own site cited as a source;
• being described accurately (pricing, features, audience).

Step 2: Build a prompt set

Write 10 to 30 prompts that reflect how real buyers ask. Mix five types: best-of ("best X for Y"), problem-led ("how do I stop late invoice payments"), comparison ("X vs Y"), alternatives ("alternatives to Y") and brand questions ("is X good for agencies?"). Our free buyer prompt generator can give you a starting list.

Keep the set stable. If you change prompts every week, you can't tell whether a change came from the engines or from your wording.

Step 3: Choose engines

Each engine has its own retrieval and style, so results differ. Track the engines your audience is most likely to use. Citewise AI supports ChatGPT, Gemini, Perplexity and Claude, each queried through its official API with web search enabled.

Step 4: Define your brand and competitors precisely

List your brand name, common alternative spellings and your domain. Then list the three to eight competitors you most often lose deals to. Precise names avoid false matches: if your brand is a common word, add a more distinctive variant.

Step 5: Run on a schedule

Run weekly for most brands, daily around launches or campaigns. Each run costs one credit per prompt per engine, so 15 prompts on 4 engines weekly is 60 credits a week, around 260 a month. The pricing calculator estimates this for your own setup.

Step 6: Read the results the right way

• Visibility (share of answers mentioning you) is your headline number.
• Share of voice shows how you compare with competitors in the same answers.
• Average position tells you whether you're the first recommendation or an afterthought.
• Top sources show which domains shape answers in your category: your outreach list.

Always open a few full answers. The context of a mention matters: being named as "an expensive option" is different from "the best choice for freelancers".

Step 7: Act, then verify

Pick one or two improvements based on what you saw, such as a clear comparison page, fixing a crawler block, or correcting outdated information on a site that is frequently cited. Then watch the next several runs. Real improvements show up as sustained changes, not single spikes.

Doing it by hand vs. automating

Manual tracking works for a quick check, but it gets unreliable fast: logged-in apps personalise answers, copying results into spreadsheets is slow, and you lose history. Automating with official APIs gives you consistent conditions, stored answers and trend charts. Try Citewise AI free; every account gets monthly credits.