• Dofollow Digest
  • Posts
  • Dofollow Digest #56: Microsoft's framework for getting recommended by AI

Dofollow Digest #56: Microsoft's framework for getting recommended by AI

Author

Eric Carrell
January 23, 2026

Hey, it's Eric 👋

Big picture stuff this week: Microsoft put out a 16-page guide on optimizing for AI search. It's probably the clearest breakdown I've seen from any major tech company on what actually matters when you're trying to get your SaaS mentioned by AI assistants.

We'll dig into the details below, but the short version is this: Microsoft doesn’t think SEO isn't going away. It's just getting a few new teammates.

🚨New Product: Branded Web Mentions for AI Visibility

We’re launching our new product: Branded Web Mentions in the coming weeks - we make sure high-quality sites reference your brand in a relevant way that drives sentiment & context to your brand. These are specifically tailored to help companies get more visibility in AI searches. We are opening a waitlist ahead of the public launch to prioritize SaaS and tech companies, qualify for fit, and share full details before spots open.

🔍 DEEP DIVE: Microsoft's Framework for Getting Recommended by AI

Microsoft recently published a guide called "From Discovery to Influence" that lays out how to optimize for AI search surfaces. Roger Montti broke it down over at Search Engine Journal, and there are some useful takeaways for SaaS companies thinking about visibility beyond traditional search.

The guide introduces two concepts that complement traditional SEO: AEO (Agentic Engine Optimization) and GEO (Generative Engine Optimization). AEO focuses on making your content easy for AI assistants to retrieve and present as direct answers. GEO is about clarity, trustworthiness, and authoritativeness inside generative AI systems.

What I found most interesting is how Microsoft frames the competitive landscape. They describe it as a shift from "discovery" to "influence." The way they put it: SEO helps your product get found. AEO helps AI explain it clearly. GEO helps AI trust it enough to recommend it.

That's a useful mental model.

The guide also breaks down how AI assistants actually process product recommendations. When someone asks Copilot for a suggestion, it pulls from three data layers: crawled web content (your brand positioning and general knowledge), product feeds and APIs (current prices, availability, specs), and live website data (reviews, promotions, delivery estimates). The AI aggregates all of this before making a recommendation.

Microsoft's three-part action plan is practical. First, get your technical foundations right: structured data, schema markup, and alignment between what's on your page and what's in your feeds. Second, optimize content for intent and clarity by writing descriptions that lead with benefits, using headings that match how people ask questions, and adding modular content blocks like FAQs and specs. Third, build trust signals through verified reviews, real-world validation like press and certifications, and consistent, factual claims.

What this means for SaaS companies:

The most interesting part of Microsoft's framework is that three data layers thing.

AI assistants aren't just pulling from your blog posts and landing pages. They're also looking at product feeds, APIs, and live website data like reviews, pricing, and delivery info.

For SaaS companies, this shifts the conversation. Your pricing page structure matters. Your G2 and Capterra reviews matter. Whether your feature specs are machine-readable matters. The accuracy of your integration documentation matters.

Most SaaS SEO conversations focus on content: blog strategy, keyword targeting, thought leadership. That's still important, but Microsoft's guide suggests AI recommendations are being shaped by a lot of stuff that traditionally lived outside the SEO conversation. Schema markup on your product pages. Consistency between what's on your site and what's in your data feeds. Whether your claims are verifiable.

This probably means SEO and product teams need to talk more. If your pricing page is a static image or your feature list is buried in unstructured paragraphs, that's not just a UX issue anymore. It's potentially an AI visibility issue.

The good news: SaaS companies already have most of this data. Integrations, features, pricing tiers, customer reviews. The work is making it structured and accessible, not creating it from scratch.

🔗 LINK ROUNDUP

Til next time,

Eric