Skip to content

Either you’re in the answer, or you’re not.

Buyers stopped scrolling ten blue links. They ask, and something answers. Findability OS is the system for being the business that answer names — and the instruments that tell you whether it does.

Findability OS Scott Sumner
The Scoreboard Awaiting input

Answer engines don’t rank you — they either name you or they don’t. This builds the exact question set to find out, phrased the way your buyers actually ask.

Ranking was a position. Being recommended is a judgement — made about you, without you in the room.

Instrument one

Are you in the answer?

The Scoreboard is a fixed set of real buyer questions, asked in fresh sessions across every answer engine your market uses, scored for presence, accuracy, and how you get described.

It is deliberately boring and deliberately repeatable. A fixed instrument you re-run monthly beats a clever one-off audit, because the only thing that proves the work is the score moving.

Nobody can guarantee placement in an answer engine, and anyone promising it is selling something. What you control is the source material: plain pages, consistent facts, quotable proof, full crawlability.

Scored across
ChatGPTOpenAI
ClaudeAnthropic
PerplexityPerplexity
GeminiGoogle
AI OverviewsGoogle
CopilotMicrosoft
Instrument two

Is it sending anyone?

TrafficWatch runs at the edge, in the request path, in front of your site — not as a script on the page. It sees every request: who arrived from an AI assistant, which company they came from, and which AI crawlers are reading you.

That last part matters more than it sounds. Analytics that run in the browser cannot see a crawler at all, because crawlers don’t run JavaScript. The machines deciding whether to recommend you are invisible to the tool most businesses trust.

Managed for you. Nothing to install, no tag to paste, no weight added to the page.

Crawlers it watches for
GPTBot · OAI-SearchBotOpenAI
ClaudeBot · Claude-UserAnthropic
PerplexityBotPerplexity
Google-ExtendedGoogle
Applebot-ExtendedApple
Meta-ExternalAgentMeta
BytespiderByteDance
CCBotCommon Crawl
 Page scriptAt the edge
Sees AI crawlers No — bots run no JS Yes
Survives ad blockers No Yes
Weight added to page A script None
Catches a two-second bounce Often not Yes
The system

Findability OS

One framework, four things a machine has to be able to do with your business before it will put you in front of a buyer. The instruments measure it. The audit tells you which part is broken.

Find Crawlable, complete, and actually reachable by the agents doing the reading.
Understand Plain pages and structured data that agree with each other about what you do.
Trust Proof a machine can quote, and facts that stay consistent everywhere you appear.
Recommend Enough context for a system to know when you belong in the shortlist.
From the blog

Writing on AI search.

The method gets worked out here first — what changed in AI search, what it means for a business that already has customers, and what to do about it.

These pages are also the source material. If an answer engine is going to describe what I do, this is what it reads.

Latest
  1. SEO is not dead. Search behavior is changing.
  2. AI Where the Drag Is Obvious
  3. 10 AI Terms Business Owners Need to Know
  4. AI Is Not the Internet Again
  5. Stop Prompting AI Like a Search Box. Start Building Loops.
  6. AI Can Open the Door. But Who Watches the Lobby?
Where people start

The Findability OS Audit

Not a free automated scan. A strategic read of what machines see when they look you up, ending in a fix order — what matters now, what can wait, and what will actually move the number.

For established businesses with something to lose. Findability multiplies what already exists; it cannot multiply zero.

What gets reviewed
  • Positioning clarity — can a buyer tell what you do in one pass?
  • Website structure, crawlability and conversion paths
  • Technical findability — schema gaps, thin pages, broken basics
  • AI search testing against real buyer prompts
  • Google visibility, branded search and entity signals
  • Proof and trust signals a machine can actually quote
  • A competitive snapshot against three rivals
  • A fix priority map, in order