Why I Built a Tool That Roasts Your Startup's Positioning

Most startup homepages sound identical.

I don’t mean similar. I mean if you stripped the logos off the top 50 results for "AI platform for teams," you could not tell them apart. Same hero headline structure. Same vague value prop. Same "trusted by" badge row. Same nothing.

I know this because I spend a disproportionate amount of my week reviewing startup messaging. On positioning calls, in teardowns for clients, in late-night rabbit holes where I open 30 competitor homepages and screenshot the ones that actually say something. The screenshot folder is always thin.

So I built a tool that automates the brutal first impression. The Positioning Roaster reads your homepage copy and tells you, in 10 seconds, whether your messaging would survive contact with a real buyer. It scores you, roasts you, and gives you a fix list.

Here’s why I built it that way.

The Problem This Solves

Founders don’t realize their messaging is generic until someone tells them.

That’s not a criticism. It’s a structural blind spot. When you’ve spent months building a product, you’re so close to it that everything feels specific. You know the nuance. You know what makes your approach different. The problem is that none of that nuance made it to the homepage.

What made it to the homepage is the version that got approved by three co-founders, reviewed by an advisor, softened to avoid offending any potential buyer persona, and ultimately compressed into something that could apply to literally any company in the category.

I was having the same conversation on every positioning call: "Your copy says you do X for Y, but so does everyone else. What do you actually do differently?" And then we’d spend 45 minutes excavating the real positioning that was hiding underneath the safe version.

The roaster automates that first question. It’s the brutal first impression you’re not getting from your team, your investors, or your mom.

How the Scoring Works

The tool evaluates your positioning across four dimensions:

Clarity — Can a stranger understand what you do in 5 seconds? Not what category you’re in. What you do. For whom. With what outcome. Most startups score high on category and low on specifics, which means a visitor knows you’re an "AI platform" but has no idea why they should care.

Specificity — Does your copy name a real audience, a real pain, or a real outcome? Or is it written for "businesses" who want to "streamline operations"? I’ve written about this before — specificity is the sharpest weapon in positioning. The more specific your language, the fewer people it applies to, and paradoxically, the more it converts.

Differentiation — If I pasted your headline onto a competitor’s site, would anyone notice? This is the dimension most startups fail. Not because they’re not different, but because their copy doesn’t reflect the difference. They write for the category instead of writing for the wedge.

Value Clarity — Does the visitor know what they get? Not features. Outcomes. "Automates your invoice processing" is a feature. "Close your books in 2 days instead of 12" is value clarity. The gap between those two statements is the gap between a bounce and a signup.

These four dimensions aren’t arbitrary. They’re the lenses I use in every positioning engagement. If you score above 70 on all four, your messaging is in fighting shape. If any single dimension drops below 40, you have a positioning leak that’s costing you conversions whether you know it or not.

What the AI Actually Evaluates

The roaster reads exactly two things: your headline and your meta description.

That’s it. Not your features page. Not your pricing. Not your about section. Just the copy a buyer sees in the first 5 seconds — the positioning surface area that matters most.

The first 5 seconds of your homepage carry more weight than the entire rest of the site combined.

Share on X

Why? Because buyers don’t read your website. They scan your homepage, decide if you’re relevant, and either stay or bounce. The average time-on-page for a startup homepage is under 30 seconds. If your headline doesn’t land in that window, nothing else you’ve built matters.

This is why I intentionally limited the tool’s input. I don’t want it to give you credit for having a good features page or a compelling case study buried three clicks deep. I want it to answer the only question that matters: does your homepage earn the next click?

Why Roasting Works Better Than a Checklist

I could have built this as a positioning checklist. Score yourself 1–5 on clarity. Rate your differentiation. Fill out this worksheet.

Nobody would have used it.

Checklists get bookmarked and forgotten. They’re homework. A roast is an event. A score plus a roast line is memorable and shareable. Founders actually send their results to co-founders. They post them on X. They screenshot the scorecard and put it in Slack.

The roast is the distribution mechanism. It’s the thing that makes someone engage with positioning feedback they would have otherwise ignored. I’ve seen founders get a score of 34 and immediately rewrite their homepage — not because a consultant told them to, but because the roast made it feel urgent and real.

That’s the psychology. Positioning advice is valuable but boring. Positioning feedback with a specific score and a specific burn is neither boring nor ignorable.

The Real Goal

The roast is the hook. The improvement plan is the payload.

Every result comes with three to five actionable fixes — not generic advice like "be more specific," but targeted suggestions based on what the AI actually read in your copy. It tells you what’s weak, why it’s weak, and what a stronger version would look like.

The goal isn’t humiliation. It’s clarity. Most AI startups will die with great products because they never figured out how to communicate what makes them worth choosing. The roaster is designed to compress that realization into 10 seconds so founders can actually do something about it.

If you score below 65, the tool suggests booking a positioning call. Not as a hard sell — as a triage response. If your homepage is leaking trust at that level, a rewrite from a checklist isn’t going to cut it. You need someone to help you find the angle.

If you score above 65, it points you to the headline grader to keep sharpening. Because good positioning isn’t a one-time event. It’s an ongoing refinement.

Try It

Roast your startup’s positioning. It takes 10 seconds, it’s free, and it’s more honest than anyone on your team will be.

If you want the full positioning framework behind every score, start with how to position your AI startup when everything sounds the same. That’s the long version of what the roaster does in a flash.

- MM