LLMs are sending us leads, and we didn't pay for ads

Something Unexpected Started Happening

Something specific has been happening at BlackBox Vision. At first it felt like an anomaly; now it looks like a pattern.

Since the start of the year, we've been detecting a new qualified lead acquisition channel that wasn't on our radar: LLMs.

We have now counted 20 qualified leads in three months who, when asked how they found us, gave the same answer: "ChatGPT." Sometimes it was Gemini. Sometimes Claude. In a few cases, the model even shared our direct contact details.

It was not magic. Last year, Manu, the team, and I started to see this discovery shift coming.

We Saw the Shift Coming

We knew people were changing how they search for products and services. The move from Google to answer engines was accelerating, and we decided to redesign our website with a completely different approach.

Aesthetics mattered, but they were not the center of the work. Semantic clarity was.

We wanted our value proposition to be easy to understand at a structural level. It could not just look good; it had to be machine-readable. We focused on getting the content relationships right and nailed the technical foundations: strict HTML usage, meta tags, URL naming, sitemap, robots configuration, and structured data.

Today many call this AEO (Answer Engine Optimization). For us, it's just SEO with a strategic and technical twist.

What Changed in Practice

The shift was not about adding fancy product features. It was about discipline:

• Direct content: no vague marketing language. We explain what we do, who we help, and which problems we can credibly solve.
• Technical hygiene: Proper HTML semantics, structured data, clean URLs, complete meta tags. The boring stuff that machines actually read.
• Intentional interlinking: Every page connects to related content in a way that helps both humans and AI understand the full picture.
• Openness to AI crawlers: We made sure our robots.txt and sitemap welcome the models we want recommending us.

None of this required a massive investment. It required focus and the willingness to prioritize substance over surface.

Why This Matters for Startups

If you're a startup founder or CTO, this should be on your radar. The lead acquisition landscape is fragmenting. People are asking AI models for recommendations before they ever type a Google query. And unlike paid channels, you can't buy your way into an LLM's answer. You earn it.

The competitive advantage here is timing. Most companies haven't adapted their web presence for AI discovery yet. The ones that move early will compound that advantage as LLM usage grows.

Here is why this channel matters:

• Zero acquisition cost: No ad spend required
• High intent leads: People asking LLMs for solutions are actively looking to buy
• Trust transfer: A recommendation from ChatGPT or Claude carries implicit credibility
• Compounding returns: Once a model learns to recommend you, it tends to keep doing so

The Lesson

Seeing those changes we implemented months ago now converting into qualified leads confirms the effort was worth it.

There is real value in getting ahead of how machines read your business. The method is not secret; most companies simply have not started yet.

The question isn't whether AI models will influence how people discover services. They already do. The question is whether your business is positioned to be recommended.