Why look at topics, not only keywords?
AI engines do not answer with a list of pages that match a phrase. They answer with what they know about a subject and the brands they connect with it. When someone asks about fitted kitchens, the answer might mention worktop materials, design visits, guarantees and made to measure units, and name the firms it links with each.
If the engines link those things to your competitors and not to you, you are less likely to be named. A keyword report will not show this, because the gap is in what the engines know, not in a single phrase. Topic gaps shows it directly.
How does a topic check work?
You add a topic, such as a service line, and choose up to three competitors. The competitors you already track in your prompts are filled in for you. Then each check runs in four stages.
- Ask the engines. Each engine in your plan is asked what it knows about your brand in relation to the topic, and the same for each competitor. One neutral question asks what matters most about the topic itself.
- Break the answers into sub-topics. The answers are read and turned into one shared list of sub-topics, such as products, features, standards, guarantees, places and types of customer. For each answer, the platform records which sub-topics it connects with that brand.
- Check your website. The pages on your site that best match the topic and each sub-topic are read, and each sub-topic is marked as covered, only mentioned, or not found.
- Score and suggest fixes. Each brand gets a knowledge score. Gaps are ranked by how strongly the engines link them to competitors, and each one gets a short suggested fix.
A check usually takes two to four minutes. You can see every answer in full, so nothing is hidden behind the score.
What do you see?
- Knowledge scores. Your score and each competitor's, out of 100, with a trend line once a topic has been checked more than once.
- A short summary. Two or three sentences on the overall picture for the topic.
- Gaps. Each one shows the engines that connect it with you and with each competitor, whether your site covers it, how serious it is, and what to do.
- Strengths. Sub-topics at least half the engines already connect with you, with the page that covers each one.
- A topic map. Every sub-topic against every brand and engine, plus what your site covers, in one table.
- The answers. Each engine's full answer for each brand, with the pages it cited.
What kinds of gap are there?
| Gap | What it means | The usual fix |
|---|---|---|
| Not found on your site | Engines link it to competitors, and the pages checked do not cover it | Add a section or page, if you offer it |
| Only mentioned on your site | It is named in passing, with no clear section | Give it a heading and specific detail |
| On your site, not picked up | Your site covers it, but engines still do not link it to you | Make it easier to find, link to it and get it mentioned elsewhere |
Gaps are marked critical when the engines link a sub-topic to a competitor strongly and not to you at all. Significant and minor gaps follow. Anything that clearly does not fit your business is kept apart in a list marked as probably not relevant.
What happens after a gap is found?
Each gap can go to Work as a task with one click. The task carries the suggested fix, the page to update and a short checklist. When the topic is checked again, you can see whether the engines now link that sub-topic with you.
Topic gaps sits well next to content opportunities, which starts from your Search Console data and rankings, and page optimisation, which checks each page in detail. Our guide to how AI engines choose sources explains why clear, specific pages tend to be picked up.
Where it fits
Topic gaps is one of the Improve tools. It works alongside AI visibility tracking, which measures whether you are named for the questions buyers ask, and competitor share of voice, which shows who is named instead. Plans and credit prices are on the pricing page.