I Asked Perplexity Who Audits SaaS Emails. My Own Tool Answered Back.

TL;DR

A Different Question

Last time I ran this test, the question was about ToolIndex, a directory. Directories are easy for AI models to compare — they’re all doing the same job with different numbers attached.

This time the question was about Strategic Flow itself, the actual diagnostic underneath everything else I’ve built. Harder thing to get an AI model to describe correctly, because it’s not a list of numbers. It’s a method.

So I typed the question a SaaS founder would actually ask — not my product name, just the problem: “best tools for auditing SaaS product emails for structural and conversion issues.”

The Category I Didn’t Choose

The answer came back sorted into three buckets: message architecture audit tools, lifecycle audit tools, and deliverability scanners. Strategic Flow was named directly in the first bucket, described as the fit for “structural issues after the open.”

I didn’t write that category name anywhere. Nobody at Strategic Flow chose the phrase “message architecture audit.” The model built that label on its own, from whatever it had read, and decided that’s the shelf my product belongs on.

That’s the part worth sitting with. You can optimize a page all day. You can’t fully control the category a model decides to file you under. You can only make the correct category easy enough to infer that it gets it right.

Who Else Was in the Room

Strategic Flow wasn’t alone in the answer. A tool called EmailFunnelAI showed up too, sorted into the lifecycle bucket — the one that scores onboarding, activation, and retention as a system rather than a single email.

Different shelf. Not a competitor for the same query, a neighbor. If anything that’s the useful outcome: the model isn’t confusing the two products, it’s drawing the boundary between “audit one email’s structure” and “audit the whole lifecycle” correctly, and putting each tool on the right side of it.

Running It Twice

One clean answer could be luck. So I ran the identical query again, fresh session, no memory of the first run carried over.

Same category. Same placement. Same framing — structural fit for post-open conversion problems.

Consistent, not a one-off. That’s the difference between a lucky citation and an actual position in the model’s understanding of the category.

What This Actually Proves

Not much, yet. Two identical queries in one search engine is a signal, not a trend line. The real test is whether the placement holds when the question is asked without the word that makes it easy — without “SaaS” attached, from a different angle, on a different model entirely.

What it does prove is narrower and more useful than “we’re famous now”: the language on the site is specific enough that a model extracted the right shelf without being told. That’s the actual goal of writing for AI search — not keyword density, legibility.

Check Where You Stand

Same close as last time, because the advice didn’t change. Ask an AI model what it thinks your product is, in the words a buyer would actually use, not your own product name. Do it twice. See if the answer holds.

If it doesn’t name you at all, that’s chapter one. If it names you but files you under the wrong shelf, that’s a different fix entirely — one about clarity, not visibility.