Should I hire a GEO agency or do it in-house? Too many agencies claiming expertise now

I helped build the GEO practice at a major agency. Let me share the honest reality:

The truth about GEO agencies:

• Most are repurposed SEO/content agencies
• Few have genuine AI visibility expertise
• The field is too new for “5 years experience”
• Many are selling what they’re still learning

Questions that separate real from fake:

1. “How do you track AI visibility?”

   • Good answer: “We use Am I Cited/similar tools, track citations across platforms, measure share of voice”
   • Red flag: “We have proprietary methods” or can’t name specific tools

2. “Can you explain how AI systems decide what to cite?”

   • Good answer: Explains RAG, training data, authority signals
   • Red flag: Vague generalities or “it’s complex”

3. “Show me a case study with AI visibility metrics”

   • Good answer: Specific numbers - citation rates, platforms, timeline
   • Red flag: Only traditional SEO metrics or “we can’t share due to NDAs”

4. “What’s the timeline for results?”

   • Good answer: “6-12 months to see meaningful AI citation improvement”
   • Red flag: “Results in 30-60 days”

Pricing reality: At $10K/month, you should expect:

• Dedicated strategist
• Regular AI visibility reporting
• Content optimization recommendations
• Citation tracking and competitive analysis
• Not just content production

Counter-perspective: We chose in-house over agency, and I’d do it again.

Why we went in-house:

• GEO is too new for “experts” to exist
• Our SEO person could learn this
• Long-term capability matters more than short-term speed
• Agencies can’t replace domain expertise

What we invested instead of agency fees:

• Am I Cited subscription: $300/month
• Training/courses: $2K one-time
• Time allocation: 25% of SEO person’s role
• My involvement: 5 hrs/month for strategy

Total cost: ~$4K/month vs $10K agency

Results after 8 months:

• AI visibility up 35%
• In-house expertise built
• No agency dependency
• Can iterate faster

When agency DOES make sense:

• Zero internal capability
• No bandwidth to learn
• Need results for a specific deadline
• Want to accelerate learning (hire to teach)

When in-house is better:

• Have SEO foundation to build on
• Can allocate 20-25% of someone’s time
• Care about long-term capability
• Budget-conscious

Let me list every red flag I’ve seen in GEO agency pitches:

Credibility red flags:

• “5+ years of GEO experience” (impossible)
• Can’t explain how LLMs work at basic level
• No mention of specific tracking tools
• All case studies are traditional SEO metrics
• “Proprietary AI algorithm” for optimization

Process red flags:

• Guaranteed citation rates or rankings
• Results promised in < 90 days
• Focus on content volume over strategy
• No competitive analysis offering
• Can’t explain their measurement approach

Contract red flags:

• Require 12+ month commitment upfront
• Own your content or data
• Won’t share raw visibility data
• Hidden fees for reporting or tools
• No performance benchmarks

What legitimate agencies do:

• Admit GEO is evolving and they’re learning too
• Use standard tools and explain them
• Show realistic timelines (6-12 months)
• Provide raw data access
• Offer flexible engagement models
• Can teach, not just do

If an agency hits more than 2 red flags, walk away.

I own an agency that now offers GEO. Let me be honest about the industry:

What we actually know:

• How to track AI visibility (tools exist)
• Content structure principles for AI
• General patterns in what gets cited
• How to measure and iterate

What we’re still learning:

• Why some content gets cited over similar content
• Platform-specific nuances
• How often citation patterns change
• Long-term optimization strategies

How honest agencies should position: “We have strong methodology and tools. We’ll track aggressively, learn from data, and iterate. We don’t have all the answers because nobody does yet.”

Why clients should care: The agencies claiming certainty are the ones to avoid. This field is too new for certainty.

My recommendation for $10K budget: Consider splitting:

• $3K/month: Tools + in-house time allocation
• $3K/month: Agency consulting (not execution)
• $4K/month: Reserve for content/execution

This way you build capability while getting expert guidance.

Here’s the evaluation framework I use when buying GEO services:

Scoring matrix (0-5 each):

Deal breakers (automatic disqualification):

• Can’t provide AI-specific case study
• Won’t give you raw data access
• Require > 6 month initial commitment
• Can’t name specific tools they use
• Promise guaranteed results

Good signs:

• Admit limitations of current knowledge
• Propose pilot/testing approach
• Use industry-standard tools
• Offer training component
• Flexible contract terms

Reference check questions:

• “Did they deliver measurable AI visibility improvement?”
• “How long before you saw results?”
• “What would you do differently?”
• “Would you recommend them specifically for GEO?”

We did a hybrid approach that worked well:

Phase 1 (Months 1-3): Agency consulting

• Hired agency for strategy only ($5K/month)
• They built our playbook and trained our team
• Set up tracking and measurement
• No content execution

Phase 2 (Months 4-6): Execution with support

• Reduced to $2K/month advisory
• Our team executed the strategy
• Agency reviewed and course-corrected
• Built internal confidence

Phase 3 (Months 7+): Mostly in-house

• $1K/month check-ins
• Full in-house execution
• Agency for complex challenges only
• Strong internal capability

Total agency spend: $24K over 7 months

What we have now:

• In-house capability
• No ongoing dependency
• Can scale without agency constraint
• Saved vs. $10K/month ongoing

The key insight: The right agency teaches you to not need them. Be suspicious of agencies that create dependency.

Whatever you decide, verify independently:

Get your own Am I Cited account (or similar).

Why this matters:

• Agencies can cherry-pick favorable data
• You see continuous data, not monthly snapshots
• Can compare performance after agency ends
• Keeps agency accountable

What to track yourself:

• AI citation rate by platform
• Share of voice vs competitors
• Specific URLs/content getting cited
• Trend over time

Story from experience: Agency showed us “45% improvement in AI visibility.” Our independent tracking showed 12%.

The difference? They measured different queries, on different platforms, at convenient times.

The rule: If agency won’t give raw data access or objects to independent tracking, they’re hiding something.

Good agencies welcome verification. They know their work stands up to scrutiny.

This thread saved me from making a bad decision. Here’s my revised approach:

What I was going to do: Sign a 12-month contract with the agency that had the best pitch.

What I’m actually going to do:

Month 1: Foundation

• Sign up for Am I Cited (independent tracking)
• Baseline our current AI visibility
• Interview 3 agencies using the evaluation framework

Months 2-3: Consulting engagement

• Hire top agency for consulting only ($5K/month)
• Focus on strategy, playbook, team training
• No execution commitment yet

Month 4: Decision point

• Evaluate: Can we execute in-house?
• If yes: Reduce to advisory relationship
• If no: Consider execution support

Questions I’ll ask every agency:

1. How do you track AI visibility specifically?
2. Show me an AI-specific case study with metrics
3. What will you teach our team?
4. What’s your honest timeline for results?
5. Can we have raw data access?

Red flags I’m watching for:

• Claims of 5+ years experience
• Guaranteed results
• 12-month commitment pressure
• “Proprietary algorithms”
• Won’t give data access

Thanks everyone. Much better equipped now.