KnitKnot
You are reading the agent-optimized layer of this page: the literal markdown we serve to AI crawlers and assistants, shipped in the page source of every visit. KnitKnot finds the claims AI repeats about a company, traces them to their sources, and helps the company correct the record.

# Getting started with KnitKnot

What KnitKnot does, how the benchmark → report → fix loop works, and what to expect in your first week.

KnitKnot measures how AI models : ChatGPT, Claude, Perplexity, and Gemini : represent your company when buyers compare vendors. It runs real buyer questions through each engine, scores every response at the claim level, and turns the gaps into prioritized playbooks.

Everything in the product hangs off one loop:

  1. 1. **Benchmark** : run your question library across the four engines.
  2. 2. **Report** : see your AI Presence Score, where you win and lose head-to-head, and which sources the AIs are citing.
  3. 3. **Fix** : work through the issues and playbooks KnitKnot generates from the gaps.
  4. 4. **Re-run** : benchmark again and watch the score move as your fixes get picked up.

## Sign in

Go to [app.knitknot.ai](https://app.knitknot.ai/signin) and sign in with your work email. Accounts are organized by company domain : teammates with the same email domain join the same account, and your company's data lives in a shared workspace.

If your company doesn't have a workspace yet, [request a benchmark](https://knitknot.ai/) and we'll set one up : the first benchmark is free.

## What's already set up for you

By the time you sign in, your workspace usually already has:

  • - **A researched benchmark model** : your capabilities, positioning, products, buyer roles, and verified facts, each backed by evidence.
  • - **A competitor set** : the vendors AI actually compares you against, each with the same depth of profile.
  • - **A question library** : buyer questions grounded in real Google search queries with monthly volume data, not synthetic templates. Questions persist across runs, so every benchmark measures the same questions and results stay comparable over time.

You can review and edit all three from the console : see [Run your first benchmark](/docs/run-your-first-benchmark/) for where each lives.

## Your first week

  • - **Day 1** : read your report ([how to read it](/docs/read-your-report/)), skim the answers behind the headline numbers, and sanity-check the competitor set.
  • - **Day 1–2** : review the open issues: factual errors, missed capabilities, and losing comparisons, ranked by severity.
  • - **Week 1** : pick one playbook and ship it. Each one maps linked evidence to a create, revise, or repair brief.
  • - **Ongoing** : benchmarks re-run on a schedule; **Overview** shows the score trend and work that needs attention.

Raw mirror of this content: https://knitknot.ai/docs/quickstart.md. Site-wide summary: /llms.txt · full content: /llms-full.txt

Docs navigation
Docs › Getting started

Getting started with KnitKnot

What KnitKnot does, how the benchmark → report → fix loop works, and what to expect in your first week.

Updated

KnitKnot measures how AI models : ChatGPT, Claude, Perplexity, and Gemini : represent your company when buyers compare vendors. It runs real buyer questions through each engine, scores every response at the claim level, and turns the gaps into prioritized playbooks.

Everything in the product hangs off one loop:

  1. Benchmark : run your question library across the four engines.
  2. Report : see your AI Presence Score, where you win and lose head-to-head, and which sources the AIs are citing.
  3. Fix : work through the issues and playbooks KnitKnot generates from the gaps.
  4. Re-run : benchmark again and watch the score move as your fixes get picked up.

Sign in

Go to app.knitknot.ai and sign in with your work email. Accounts are organized by company domain : teammates with the same email domain join the same account, and your company’s data lives in a shared workspace.

If your company doesn’t have a workspace yet, request a benchmark and we’ll set one up : the first benchmark is free.

What’s already set up for you

By the time you sign in, your workspace usually already has:

  • A researched benchmark model : your capabilities, positioning, products, buyer roles, and verified facts, each backed by evidence.
  • A competitor set : the vendors AI actually compares you against, each with the same depth of profile.
  • A question library : buyer questions grounded in real Google search queries with monthly volume data, not synthetic templates. Questions persist across runs, so every benchmark measures the same questions and results stay comparable over time.

You can review and edit all three from the console : see Run your first benchmark for where each lives.

Your first week

  • Day 1 : read your report (how to read it), skim the answers behind the headline numbers, and sanity-check the competitor set.
  • Day 1–2 : review the open issues: factual errors, missed capabilities, and losing comparisons, ranked by severity.
  • Week 1 : pick one playbook and ship it. Each one maps linked evidence to a create, revise, or repair brief.
  • Ongoing : benchmarks re-run on a schedule; Overview shows the score trend and work that needs attention.