Building case studies

A case study documents a customer's situation, verified work, and observed results. Get permission to share the evidence and state the limits of any before-and-after comparison.

The case study template

Every GEO case study follows the same structure:

1. the setup (2-3 sentences)

Who is the client? What industry? How big? What was their situation?

"[Client Name] is a [industry] company serving [audience]. Despite strong organic search rankings, they were virtually invisible to AI models: scoring just [X]/100 on their initial AI Visibility Audit."

2. the problem (1 paragraph)

What specific challenge did they face? Use the audit data.

"When potential customers asked ChatGPT, Perplexity, or Google's AI Overview for recommendations in [category], [Client Name] wasn't mentioned. Their top competitor appeared in 68% of AI responses while they appeared in just 3%."

3. the approach (2-3 paragraphs)

What did you do? Be specific about tactics without revealing proprietary methodology.

Cover:

Which pillars you focused on and why
Key actions taken (citation volume, entity fixes, press campaigns)
Timeline and phases

4. the results (this is the money section)

Hard numbers, before/after comparisons, specific metrics:

AI Visibility Score: X → Y (+Z%)
Share of Model: X% → Y% across N models
Citations placed: N threads across N platforms
Specific AI mentions: "Now appears in Perplexity's top 3 recommendations for [keyword]"
Business impact if available: website traffic from AI referrals, leads attributed

5. the quote (1-2 sentences)

A client testimonial that reinforces the results in their own words.

6. key takeaway (1 sentence)

What can the reader learn from this case?

What metrics to highlight

Prioritize metrics that prospects can understand without GEO expertise:

Must include

AI Visibility Score change (before → after)
Share of Model percentage change
Timeline to results
Competitor comparison shift

Include if available

Website traffic from AI referrals
Leads or sales attributed to AI visibility
Brand search volume increase (a secondary signal of AI visibility)

Skip

Internal process metrics (threads discovered, responses generated)
Technical details (SERP positions, content hashes)
Credit costs or operational details

Getting client permission

Before publishing any case study:

Ask early. Mention during onboarding that you'd love to feature them as a case study if results are strong. This plants the seed.
Show the draft. Never publish without client review. Send them the draft and let them edit their quote, adjust numbers, or request anonymization.
Offer anonymization. Some clients don't want to be named. "[A leading SaaS company in the HR space]" still works as a case study: the numbers tell the story.
Get written approval. An email confirmation is sufficient. Keep it on file.

Publishing for your own AI visibility

Here's the meta play: publishing case studies improves your own agency's AI visibility. When AI models are asked "What agencies do GEO?" or "Who offers AI visibility services?", published case studies on your website create the citation and authority signals that get you recommended.

Where to publish

Your agency website (dedicated case study page)
Your blog (detailed writeup)
LinkedIn (summary with key metrics)
Industry publications (guest post format)

SEO + GEO optimization

Target keywords like "GEO case study," "AI visibility results," "[industry] GEO success"
Add schema markup to the case study page
Reference in your own llms.txt file
Share in relevant industry forums and communities

Case study volume

Aim for:

1 case study per quarter minimum
1 per month if you have enough clients with strong results
Different industries to show breadth
Different starting points to show you can help various scenarios

A library of 4-6 case studies across different industries is the credibility tipping point where prospects stop wondering "does this work?" and start asking "how do we start?"

In Outrigger

Use the Reports module to pull the data for your case studies. Historical audit snapshots, SOM trends, and action logs provide all the numbers you need.