SUEDESCAN

Why AI recommends your competitors, and how to fix it.

The Screenshot is the thinking behind Suede Scan, written down as a book. Eleven chapters on why buyers now get their shortlists from ChatGPT, Perplexity, and Gemini, and the exact checks and repairs that decide whether you are in those answers. Written by Jason Colapietro, the operator who runs every scan, from the same runbooks the paid work uses.

Every chapter in Part I ends with a prompt you paste into ChatGPT to see your own category's answer. The book works because you watch your own gap appear, not because anyone describes it to you.

51 KB EPUB, opens in Apple Books, Kindle apps, and any reader. Also available as a PDF or as plain markdown.
What's inside

Half the book is stakes. Half is repairs.

Part I shows you where the buyers went and what absence from AI answers costs. Part II is the founder's edition of the professional audit: access, engines, structure, evidence, and a monthly measuring rhythm.

Part I · The stakes
01

The Screenshot

The one artifact nobody argues with: a machine, asked a buyer's question, naming your competitors and not you.

02

The Quiet Migration

Buying research moved from ten blue links to one synthesized answer, and your dashboards cannot see it.

03

Invisible Is the New Page Two

AI answers are named or not named. There is no position fourteen, and no alert when you are absent.

04

Ranked vs. Cited

Engines quote sources, they do not rank pages. Why SEO success has not protected you, and why that is good news.

05

Compounding Absence

Three loops that make waiting the one indefensible strategy.

Part II · The fix
06

Can the Machines Even Read You?

GPTBot, PerplexityBot, ClaudeBot, Google-Extended: the per-bot access check, by hand, in fifteen minutes.

07

The Six Engines and Who Each One Trusts

ChatGPT, Perplexity, Gemini, AI Overviews, Copilot, Claude: how each selects sources.

08

Extractable or Invisible

AI systems extract passages, not pages. The extraction test, the blocks engines lift, schema, and the honest status of llms.txt.

09

Receipts Beat Claims

Named authors, dated numbers, disclosures. Machines evaluate trust the way skeptical buyers do.

10

The Founder's Visibility Audit

All five lanes assembled into one afternoon of work, with the severity order fixes should follow.

The close
11

Measure Like an Operator

Evidence, shipped inputs, and deltas. The monthly re-scan rhythm, and why single readings are weather.

Excerpt · Chapter 1The Screenshot

If you are not in the answer, you are not losing the deal. Losing implies you competed. You were never in the room.

The Screenshot · Chapter 1 · The morning this became a business

"Traditional SEO gets you ranked. AI SEO gets you cited."

Chapter 4 epigraph

"Blocked silently, which is worse than an error."

Chapter 3 epigraph, from the operations runbook

"Never claim an outcome. Evidence, shipped inputs, and deltas only."

Chapter 11 epigraph, from the operations runbook
A
Who wrote it

The operator, not a content team.

Jason Colapietro founded Suede Labs AI and runs every Suede Scan capture personally: the prompts, the timestamped screenshots, the fixes shipped as patches and pull requests. The book is those runbooks reorganized so a non-technical founder can run the diagnosis alone. The pull quotes at chapter heads come verbatim from the working documents, including the mistakes.

B
What it does not claim

No citation promises, in the book or anywhere else.

AI answers are non-deterministic. Every answer you capture while reading is point-in-time evidence, and the book says so in its front matter, because anyone guaranteeing you a spot in an AI answer is selling something they cannot control. What you control are the inputs. The book teaches the inputs.

The trade the book makes

You get the full diagnosis method for free, forever. Suede Scan sells speed and execution to the readers who want the same work done for them, with evidence attached. That is the entire funnel, disclosed.

01Is the book actually free?
Yes. Direct EPUB download, no email, no checkout, and the full text is also readable as a PDF or plain markdown in your browser. It is the thinking behind the paid scans, published because the diagnosis should not be a secret.
02Do I need to be technical to use it?
No. Every check runs in a browser: paste prompts into ChatGPT and Perplexity, read your own robots.txt with the rules explained, view a page's source and search for your copy. The most technical artifact in the book is a checklist.
03Will reading it get my company cited by AI?
Nobody can promise that, and the book is blunt about why: engines are non-deterministic and their selection logic is not public. What it gives you is the full input-side discipline, access, structure, evidence, and entity clarity, plus a monthly measurement rhythm so you see deltas instead of guessing.
04How does the book relate to the paid Suede Scan?
Chapter 10 is the founder's edition of the audit the paid work runs. Do it yourself in an afternoon, or order it done at the services page: professional captures, prioritized fixes, refund-backed, same method as the published methodology.

Read it, then look at your own answers.

The book ends where the work starts: your category, your prompts, your screenshots. Start with the free site check, or go straight to the download.

Also as a PDF or markdown. Want the audit done for you instead? See the services.