Why SaaS buyers increasingly choose through AI
A few years ago, choosing B2B software started with Google and a dozen tabs of reviews. Today it increasingly starts with a question to ChatGPT or Claude: which platform fits this use case best, what can replace this tool, compare these three services. The AI assistant assembles an answer from several sources and names a handful of products right away. If yours is not in that answer, it does not even make the shortlist, and classic Google rankings no longer make up for it.
For SaaS this hurts for three reasons. First, technical audiences trust AI answers and use them every day at work. Second, in a new product category AI systems are only beginning to learn whom to treat as a source, and the advantage goes to whoever gets there first. Third, part of Google search is also moving to zero-click answers: AI Overviews and AI Mode show a ready conclusion, and a brand that is missing there loses both the impression and the visit.
GEO (Generative Engine Optimization) does not replace SEO, it builds on top of it. AI systems often pull sources from the top of search results, so a technically clean site remains the foundation. Beyond that, different things matter: can an exact answer be lifted from your page, is it clear what the product does and for whom, and do independent sites mention you.
Our approach, step by step
- AI visibility audit. We ask ChatGPT, Claude, Gemini, Perplexity and AI Overviews the questions your audience actually asks, and record who gets mentioned, who gets cited and in what wording. We compare your share of mentions with competitors for each topic.
- Technical foundation. We check whether AI crawlers can reach your site, documentation and tutorials, and set up robots.txt, an llms.txt file and Organization, SoftwareApplication, FAQPage and Article markup.
- Content that is easy to cite. We rework key pages into an answer format: a clear product definition up front, step-by-step instructions, use cases, comparison tables, FAQs.
- Off-site signals. We work on profiles on G2, Capterra, industry directories and specialist media, because AI systems do not trust your site alone.
- Measurement. We set up a dedicated AI assistant channel in GA4 and track sessions, sign-ups and purchases from it, along with impressions in AI Overviews and AI Mode.
- Ongoing monitoring. Every month we repeat the mention checks and show where your brand has grown and where competitors are still stronger.
Typical problems we find
- The AI channel as a black box. Traffic from ChatGPT and Claude is lumped in with referrals, and nobody knows whether it converts.
- Useful content closed to search. Tutorials or documentation carry noindex or are blocked, so AI systems have nothing to build an answer from.
- Vague positioning. The homepage says something like a platform for business growth, and AI cannot tell which problem the product solves.
- The brand trails competitors in mentions even though the product is just as good: independent sources simply write about it less.
- No structured data, so snippets and AI blocks take information from other people's pages.
Results from our projects
For an AI product with a B2B SaaS model, we opened 26 tutorials that a CMS setting had kept out of search and reworked key pages for AI answers. In the first month, impressions in AI Overviews and AI Mode grew from 818 to 3,245, visits from ChatGPT and Claude from 118 to 384, and AI Assistant channel sessions in GA4 over 28 days from 3,701 to 14,493. Part of the growth came from the client's parallel paid campaigns, and we state that openly in our reports.
For a technology company, we combined GEO with non-branded SEO and set up AI channel tracking. Within two months, sessions from ChatGPT grew from 8 to 26. The absolute numbers are still small, but the AI channel delivered its first key events, including a purchase.
For a B2B SaaS company with three domains, we prepared the ecosystem domain's content for citation in AI answers, and its AI impressions rose from 5,316 to 7,213 within a month. In another SaaS project, the audit showed the brand ranked only fourth among competitors by number of mentions in AI answers, and we identified the topics where it was mentioned less than it deserved. In a related project, technical cleanup was followed by the first referrals from AI assistants: 6 instead of zero.
For our own work we also built an SEO/GEO content platform that takes topics from Search Console queries, publishes to different CMSs and shows indexing status. We shipped the first version with 17 screens in 3 days.
Timelines and deliverables
The AI visibility audit takes the first weeks. Technical fixes such as reopening blocked tutorials can noticeably lift AI impressions in the very first month. Off-site mentions and a shift in share of mentions inside assistant answers build up more slowly, so we track them month by month. You get a map of the queries where AI should name your product, a plan of on-site and off-site changes, a configured AI channel in GA4 and a monthly report benchmarked against competitors.
Who it is for
- SaaS and B2B platforms whose buyers research solutions through AI assistants;
- products in a new category with no established leaders in AI answers yet;
- companies that already have SEO traffic but do not know what is happening in the AI channel;
- IT teams entering English- and German-speaking markets that want to be on the shortlist from day one.