Ehsan Jahandarpour
AI Growth Stack Model (AGSM)Series A

AI Growth Stack Model (AGSM): AGSM for Diagnosing Underperforming AI Tools

Using AGSM layer analysis to understand why specific AI tools are not delivering expected results.

How to Apply

1

Place every AI tool on the AGSM diagram. Identify which layer each tool operates in.

2

Rate Foundation, Execution, Optimization, and Intelligence on completeness and quality.

3

For underperforming tools, check if the layer below them scores below 7/10. If yes, the problem is the foundation, not the tool.

4

Strengthen the weakest layer first. No tool in Layer N can perform well if Layer N-1 is broken.

Expected Outcomes

  • ✓ Root cause identification for underperforming AI tools
  • ✓ Correct investment priorities (fix the foundation, not add more execution tools)
  • ✓ Improved performance from existing tools without additional spend

Real-World Examples

Common Pitfalls

⚠Blaming the tool vendor when the problem is your data infrastructure
⚠Adding more tools to a broken layer instead of fixing the existing ones

Ehsan's Insight

When a growth leader tells me "our AI tools are not working," I do not look at the tools. I look at the layer below them. Nine times out of ten, the AI tool is fine. The foundation feeding it is broken. One company spent $200K on a personalization engine and was getting a 0.3% conversion lift instead of the expected 5-15%. They were about to fire the vendor. I ran a 2-hour AGSM diagnostic. Their Foundation Layer score: 4/10. Events were misconfigured, 23% of user profiles had duplicate records, and their attribution model was last-click only. The personalization engine was making decisions based on wrong data. We spent 6 weeks fixing Foundation Layer issues (total cost: $8K in consultant time + $0 in new tools). Then the same personalization engine started delivering a 9.2% conversion lift. Same tool, same vendor, same contract. Different foundation. I have a name for this pattern: Foundation Fault Syndrome. It explains 60% of "AI tool failures."

EJ

Ehsan Jahandarpour

AI Growth Strategist & Fractional CMO

Forbes Top 20 Growth Hacker · TEDx Speaker · 716 Academic Citations · Ex-Microsoft · CMO at FirstWave (ASX:FCT) · Forbes Communications Council

Frequently Asked Questions

When should I use AI Growth Stack Model (AGSM) for stack diagnosis?
Using AGSM layer analysis to understand why specific AI tools are not delivering expected results.
What are the steps in AGSM for Diagnosing Underperforming AI Tools?
There are 4 key steps: Map tools to layers, Score each layer 1-10, Check layer dependencies, Fix bottom-up.
What results can I expect from AGSM for Diagnosing Underperforming AI Tools?
Root cause identification for underperforming AI tools. Correct investment priorities (fix the foundation, not add more execution tools). Improved performance from existing tools without additional spend.
What are common mistakes with AGSM for Diagnosing Underperforming AI Tools?
Blaming the tool vendor when the problem is your data infrastructure. Adding more tools to a broken layer instead of fixing the existing ones.