43% of B2B buyers now use AI to research software. Here is how SaaS companies can optimize their brand for AI-generated recommendations.
According to a 2025 Gartner Digital Markets survey, 43% of B2B buyers use AI assistants to research software before visiting a vendor's website. The query looks something like this:
"What's the best CRM for a 50-person remote sales team with HubSpot integration?"
Or: "Which project management tool is best for engineering teams that also need client reporting?"
These are not Google searches. They are AI prompts — and they are happening at the exact moment a buyer is forming their shortlist. If your product is not in the AI's answer, you are not on the shortlist.
SaaS products are actually well-positioned for AEO — if they optimize correctly. Here is why:
Buyer queries are highly specific and answerable. "What is the best onboarding software for teams of 50–200?" has a structured, defensible answer. AI engines love answerable questions with clear category winners.
Feature comparisons are exactly what AI engines want to cite. Structured content like "X integrates with Slack, Salesforce, and BambooHR" is precisely the kind of information AI engines extract and cite in recommendations.
Use-case pages and case studies are citation gold. A page titled "HR onboarding software for remote-first teams" with a clear answer in paragraph one is highly citable. Most SaaS sites have these pages — they just are not formatted for AI extraction.
The gap between most SaaS companies and their AEO ceiling is primarily structural. The content exists. It is not formatted for AI engines to use.
These are the five query categories that drive the most commercial value in AI-assisted research. Run these through [Prompt Performance](/features/prompt-performance) to see where you stand:
Use-case landing pages — One page per use case your product solves. Structure: title matches the query, direct answer in paragraph one, FAQPage schema with 5–8 questions about that use case.
Integration pages — Each integration your product supports should have a dedicated page with Organization + SoftwareApplication schema. Perplexity heavily cites these when buyers ask "does [tool] integrate with [platform]."
Comparison pages — "[Your product] vs [competitor]" pages are among the most cited content in AI responses. Perplexity users frequently ask for comparisons and Perplexity sources these pages directly.
FAQ page — A dedicated /faq page with 20–30 buyer-intent questions and FAQPage JSON-LD schema. This is the single page most likely to earn AI citations across all engines.
For the JSON-LD implementation across all these page types, read our [complete JSON-LD schema guide](/blog/json-ld-schema-ai-visibility).
Track two things in parallel:
MentionShare by prompt category — Group your tracked prompts by type (use-case, comparison, pricing, general category) and track your [MentionShare score](/features/mentionshare) per group. This tells you which categories you are winning and losing in AI answers.
Competitor share of voice — Use [Competitor Intelligence](/features/competitor-intelligence) to see which competitors AI engines recommend instead of you for each prompt. This directly informs your content roadmap — you are not guessing what to build, you are building what your competitors built that earned them the AI recommendation.
Week 1: Baseline scan
[Run your first scan](/sign-up) across your top 10–15 buyer-intent prompts. Record your MentionShare score for ChatGPT, Perplexity, and Gemini separately. This is your benchmark.
Week 2: Fix your top 3 missed prompts
Use the [Fix Generator](/features/fix-generator) to generate targeted FAQ blocks and JSON-LD schema for the three prompts where you have zero mentions. Publish to your homepage and primary product page.
Week 3: Add SoftwareApplication and FAQPage schema
Add SoftwareApplication JSON-LD to your product pages and FAQPage JSON-LD to your use-case pages and dedicated FAQ page. These are the schema types most associated with SaaS AEO citation improvements.
Week 4: Track progress and repeat
Re-run your baseline scan. Any prompt that now shows a mention is a win. Identify the remaining gaps and repeat the cycle. Most SaaS teams see 10–20% MentionShare improvement by the end of month one and 30–40% improvement by month three.
The SaaS companies building this practice in 2026 are creating a compounding advantage. AI engines trained to recommend them today will be harder to displace in 2027.