Sign up
All posts
Why Scooutly7 min read

Why Scooutly: the half of the picture nobody is measuring

Your rank tool sees search. Your AI tool sees answers. Neither sees why a competitor gets recommended and you don't. That gap is the product.

The short answer

Because a rank tracker tells you where you are and an AI tool tells you whether you were named — and neither one can tell you why.

Scooutly measures the visibility, then puts the competitor's price, packaging, reviews, ads and source coverage next to it. That is where the answer lives, and it is the half nobody else collects.

The rest of this piece explains how we got to that answer, and what it changes in practice.

Every company we talk to already owns a rank tracker. Many have added an AI visibility tool in the last year. Both are useful. Neither answers the question that actually gets asked in the room:

"Why do they get recommended and we don't?"

A rank tracker answers where you are. An AI visibility tool answers whether you were named. Neither one knows what your competitor charges, what they just added to their packaging, what their customers complain about, or which pages the model was reading when it named them.

That missing half is why we built Scooutly.

The problem is that visibility has two axes now

Search returns pages. AI returns an answer. They are fed by different indexes, they respond to different inputs, and they move independently. Ranking well organically while being absent from AI answers is not an edge case — it is the normal state for most brands right now.

Two axes, measured separately
Google organic74
AI Overview31
ChatGPT41
Claude36
Illustrative shape, not a measurement of any real brand. The point is the spread: one strong axis tells you nothing about the other.

If you only track the first row, you will conclude everything is fine. The three rows underneath are where the buying conversation is now happening.

Why a score alone changes nothing

Most AI visibility tools stop at a number. You are told you scored 41. You are not told why, and you cannot act on a number.

When a brand is missing from AI answers, the cause is almost always one of three things, and each has a completely different fix:

Cause What it looks like The fix
Access AI crawlers blocked in robots.txt One file change
Content shape Pages list features, don't answer questions Rewrite a few pages
Sources Third-party sites cover competitors, not you Outreach and listings

These are not the same problem. Telling someone "your score is 41" without separating them sends them to rewrite content when the real issue was a blocked crawler.

We measure the three separately, and we say which one is broken.

The part no visibility tool has

Here is the thing we kept running into: the answer to "why them and not us" is usually not about content at all. It is about positioning. And positioning lives in data that no visibility tool collects.

What each tool knows about your market
Rank trackerrank only
AI visibility tool+ mentions
Scooutly+ market
Capability comparison, not a performance claim. The bars represent how much of the question each category of tool can answer.

Scooutly puts four more things next to the visibility numbers:

Competitor pricing and packaging. We snapshot their pricing pages and diff them. When a competitor moves a feature into a higher tier or drops their entry price, you find out in the same report where your visibility dropped.

Reviews, as themes. Not a star average — the sentence that keeps repeating. A competitor's chronic complaint is often your positioning line.

Ad intelligence. Which message a competitor has been running, and for how long. A creative that has run for nine weeks is a creative that works.

Source gaps. Which sites link to and cover your competitors but not you, split into the ones you can join yourself today and the ones that need an email to an editor.

Why the method matters more than the number

Two commitments shaped the product, and both cost us features we could have shipped faster.

The score is computed, not asked. We never ask a model "rate this brand out of 100." Scores come from an open rubric applied to measured facts. The same site measured twice gets the same score. A model asked twice does not.

Every finding keeps its evidence. The raw answer, the citation, the page snapshot. If you cannot show where a number came from, nobody will act on it — and they shouldn't.

Same question, asked five times
Run 1named
Run 2absent
Run 3named
Run 4absent
Run 5named
Why a single measurement is worthless: the same question returns different answers. We ask repeatedly and report frequency — 60% here, not "yes" or "no".

If you build with Claude Code or Codex

A pattern we did not expect: the people who adopted Scooutly fastest were the ones already living inside a coding agent. The reason is mechanical. Claude Code and Codex can read your repository, your docs and your roadmap. What they cannot read is your market. Connect Scooutly over MCP and the same agent that edits your pricing page also knows what your competitor charges.

Alone+ Scooutly
Reads your code and files
Knows competitor pricing and packaging
Knows which questions you are missing from
Notices when a competitor changes something
Lists the sources you are absent from

This is a capability comparison, not a performance claim. We are not saying code written this way ranks better — we have not measured that and would not publish it if we had not. What changes is the input the agent is reasoning from.

In practice it looks like this. You ask your agent to rework the pricing page. Without Scooutly it rewrites copy based on what is already there. With Scooutly it knows a competitor moved API access into a higher tier last week, that your entry plan is now the cheapest in the category, and that "confusing pricing" is the complaint repeating in their reviews. Three facts the agent could not have guessed, and all three change what it writes.

What the agent can reason from
Your repositoryboth
Competitor pricingonly with us
Review themesonly with us
AI answer coverageonly with us
Bars show what is available as context, not how well anything performs.

The same holds for Codex. The connector is the same MCP server; only the client changes.

And it works where you already work

Scooutly runs an MCP server. Connect ChatGPT, Claude, Gemini, Grok or Perplexity once and your measurements land in the conversation you are already having — next to your roadmap, your docs, your strategy. You pay nothing extra for that; your own subscription covers the model.

If you would rather not connect anything, our assistant sits inside the dashboard and runs on credits.


The short version

A rank tracker tells you where you are. An AI tool tells you whether you were named. Scooutly tells you why — and hands you the evidence, the competitor's move, and the specific thing to fix.

That is the whole pitch. Everything else in the product is in service of being able to say it honestly.