AI-Run Operations — Catching Issues Before Customers Do
January 25, 2026
AI-Consumable Services — When AI Becomes Your Customer
January 28, 2026
This is part of the blog series, “A practical AI strategy framework — Beyond the hype”.
The first three AI strategy areas I covered — AI-powered experiences, AI-built systems, and AI-run operations — are all inward-facing. They’re about how AI helps your teams and products perform better. The next shift is different. It’s outward-facing. It’s about AI interacting with your business from the outside.
Here’s the question most companies still aren’t asking seriously enough:
When someone asks ChatGPT or another AI system for a recommendation, can that AI find you, understand you, and recommend you correctly?
AI assistants now routinely recommend products and services as part of the conversation — often without the user ever clicking a link. This is not a future scenario. It’s already happening. You have probably heard about how nearly 60% of searches never result in a click to external websites. But that’s not the full story. Zero-clicks have been slowly going up for years — with enhancements like Featured Snippets, Knowledge Panels and built-in tools (Calculator, Translator). Then, AI overview became a standard feature in Google — and other search engines. Now, for queries where an AI overview is present, the click-through rate for traditional links can plummet by up to 47%. That changes the game. It clearly tells us that the user behavior has fundamentally shifted — from searching & clicking to conversing & finding answers. Discovery is no longer about ranking higher on searches — it’s about whether AI systems can reason about you at all.
Most companies assume they’ll show up to customers because they have decent SEO, strong brand recognition, or active social channels. That assumption is risky. AI systems don’t discover companies the same way search engines do. They rely on structured information, trusted data sources, ecosystem signals, and evidence they can synthesize into an answer.
I’ve seen companies with real market share get completely bypassed in AI recommendations. Not because their product was worse — but because their information wasn’t accessible or understandable to AI systems. If AI can’t clearly infer what you do, who you’re for, and why you’re credible, you may as well not exist in that moment.
So what does AI-discoverable actually mean in practice?
First, it means having clear, structured information about what you do — your use cases, capabilities, pricing model, and customer segments — not marketing copy scattered across dozens of pages. AI systems are far better at reasoning over clarity than persuasion.
Second, it means being present in the ecosystems where AI looks for answers: developer platforms, public APIs, documentation hubs, trusted directories, and partner knowledge bases. This matters even more if you’re a software product or service. If a developer is asking Cursor or another AI coding assistant for an Email API, you need to exist in places those systems can see and trust.
Third, it means providing evidence — real customer outcomes, case studies, benchmarks, and proof points. AI recommendations increasingly lean on verifiable signals, not slogans.
Andreessen Horowitz talks about how unlike traditional search, LLMs remember, reason, and respond with personalized, multi-source synthesis. This applies even if your product isn’t “digital” in the traditional sense. AI systems reason over digital representations — structured data, interfaces, and documented behavior. Companies in traditionally non-digital categories still need to think seriously about where digital capabilities become part of their offering, because AI will reason about them whether they plan for it or not.
The companies preparing for this now are deliberately documenting their value in ways AI can understand and placing that information where AI systems are already looking. Those who don’t are quietly assuming customers will keep finding them the old way. That’s a dangerous assumption.
Next, I’ll cover AI-consumable services — where AI agents don’t just recommend you, but actually use your services to get work done. Until then, happy exploring — and may your business stay firmly on the AI navigation map as we head into uncharted space🖖.
