Skip to main content

First Week Checklist

Use the first week to make the dataset useful. You do not need hundreds of prompts. You need the right categories, competitors, and review habits.

Day 1: finish setup

1

Confirm the brand profile

Check brand name, website, logo, and competitors in Brand Profile.
2

Review onboarding prompts

Open Prompts and remove anything that is too vague, duplicated, or outside your market.
3

Add 5 to 10 high-value prompts

Cover discovery, comparison, brand, and high-intent questions. Keep them close to real buyer language.
4

Add shopping prompts if relevant

If customers buy products online, set up Shopping Prompts early so purchase intent has time to build history.

Day 2 to 3: read the first results

Start with Dashboard, then inspect the pages behind any weak cards.

Low visibility

Find prompts where the brand is not mentioned or is mentioned below competitors.

Weak sentiment

Inspect the responses and sources that shape the sentiment score.

Weak purchase intent

Check whether AI gives direct recommendations, vague mentions, or competitor-led answers.

Entity uncertainty

Verify whether AI understands the brand before relying on live web search.

Day 4 to 5: prioritize actions

Use Actions after enough analysis has completed.
1

Open Actions: Prompts

Sort underperforming prompts by metric, source, and provider. Open the detail drawer for response evidence.
2

Open Actions: Content

Run or review the content readiness audit to find pages with clarity, freshness, or entity-signal gaps.
3

Pick a small batch

Start with fixes that affect multiple prompts or high-intent shopping questions.

Day 6 to 7: create a review habit

Create a short recurring review:
  • Check the dashboard once a week
  • Add new prompts when customers, sales calls, or campaigns reveal new questions
  • Watch source changes after content updates
  • Use the competitor page after major category or messaging changes
  • Run experiments only when you have a clear hypothesis and enough prompt coverage
Do not optimize every weak prompt one by one. Prioritize sources, pages, and positioning issues that explain several weak answers at once.