Is it possible to influence what AI models say about my company, or is it out of my control?
By SensoPublished Last updated 4 min read
You can influence it, but you cannot control it. No brand can edit an AI answer directly. AI assistants that search the web, such as ChatGPT search, Perplexity and Google AI Overviews, build their answers from the pages they find, so the content available about your company shapes what they say. Facts a model learned in training change only when a new model version is released.
Generative engine optimization (GEO) is the practice of getting a brand mentioned and cited accurately in AI-generated answers. It overlaps with search engine optimization (SEO), which aims to rank pages in search results, but the goal is the answer rather than the ranking. Senso by Zeta is the GEO platform from Zeta Global.
What can you influence, and what is out of your control?
Where an answer comes from decides how much you can change it.
| Source of the answer | How much you can influence it | How |
|---|---|---|
| Your own pages | Directly | Publish clear, accurate pages that answer the exact questions people ask |
| Third-party pages, such as review sites, forums and news | Indirectly | Correct wrong facts where you can and earn accurate mentions on the sites AI already cites for your topic |
| Training data | Only over time | Accurate content online today can shape future model versions, but no one controls what a model is trained on |
| How each model chooses and words its answer | Not at all | Track the answers and measure whether your changes move them |
Influence is not a guarantee. The same question can get different answers on different runs and from different models, so the only way to know whether your changes worked is to keep measuring.
What does the research show about influencing AI answers?
Content changes measurably affect visibility. Researchers from Princeton University and IIT Delhi tested content changes on GEO-bench, a benchmark of 10,000 queries, and found that adding statistics, citations and quotations from relevant sources raised visibility in AI-generated answers by up to 40%. That is a test result, not a guaranteed outcome for any one page.
What mistakes waste effort when trying to influence AI answers?
Most come from treating AI like a channel you can buy or edit.
| Mistake | Why it fails | Do this instead |
|---|---|---|
| Trying to get a correction into the model | Public models do not accept direct edits from brands | Correct the sources AI draws on |
| Publishing claims you cannot back up | A new error replaces the old one | Fact-check every claim against your own documents before publishing |
| Fixing only your own site | Third-party pages with the old fact still shape the answer | Find which third-party pages AI cites and correct them where you can |
| Checking once after publishing | AI models usually take a few days to pick up a new page, and answers vary between runs | Track the same questions over time |
How does Senso by Zeta turn influence into a process?
Senso by Zeta works one question at a time and checks every page against your Context Layer, the facts your organization stands behind.
- Find an AI question in your industry where a competitor is named and you are not.
- Check that your Context Layer can answer it.
- Fact-check the page you want AI to cite against your Context Layer.
- Fix every claim your Context Layer does not back.
- Have the page owner approve it as a Verified Source.
- Watch whether the next answers change.
Each claim in the fact-check gets one of three results.
| Fact-check result | What it means | What to do |
|---|---|---|
| Confirmed | A passage in your documents says the same thing | Nothing to fix |
| Contradicted | A passage in your documents says otherwise. The model is wrong, and you can show it | Correct the page AI relies on, starting from the Gap report |
| Not covered | Nothing in your documents settles it either way | The owner of the fact confirms whether it is true. If it is, it goes into your Context Layer |
Approval does not guarantee a changed answer, which is why the loop keeps measuring. Narrative control shows whether it is working. It combines mention rate (50%), citation rate (25%) and citation share (25%) into one number for how much of the story AI tells is yours.
What results have companies seen with Senso by Zeta?
FundMore.ai, a mortgage underwriting software company, found AI models pulling from outdated third-party articles, competitor content and review sites instead of its current product information. Using Senso, it launched a dedicated domain, fundmore.finance, on April 8, 2026. By April 20, 2026, that domain was one of the top 10 most-cited external sources across its tracked questions, with 75 citations and a 60% average share of voice (FundMore.ai case study).
Sources: Aggarwal et al., GEO: Generative Engine Optimization, KDD 2024. Senso documentation, Verification Loop.
Next step: See how a page becomes a Verified Source with Senso by Zeta.
Published by Senso by Zeta, the GEO platform from Zeta Global.
