Who is Richard Ewing?

Richard Ewing is a Product Economist and AI Capital Auditor. He is the Founder of Exogram, a verification infrastructure for AI. He audits R&D spend and surfaces hidden capital risks like technical debt, AI cost overruns, and zombie infrastructure. He has scaled B2B SaaS from $0 to $25M ARR. Published in CIO.com, Built In, Mind the Product, and HackerNoon.

What is a Product Economist?

A Product Economist treats product decisions as economic decisions. Instead of measuring velocity or story points, a Product Economist measures Return on Invested Capital (ROIC), Cost of Goods Sold efficiency, and technical debt in dollar terms. The methodology was coined by Richard Ewing.

What is the Technical Insolvency Date?

The Technical Insolvency Date (TID) is the specific quarter where maintenance costs consume 100% of available engineering capacity, reducing feature velocity to zero. Calculate yours free at richardewing.io/tools/pdi.

Exogram is verification infrastructure for AI, founded by Richard Ewing. It prevents hallucination propagation with admissibility control planes and state-hashing commit enforcement. LLMs generate language; Exogram maintains reality.

How do I calculate technical debt cost?

Use the free Product Debt Index (PDI) calculator at richardewing.io/tools/pdi to quantify hidden technical debt in dollar terms, calculate your Technical Insolvency Date, and benchmark against industry standards.

How much does AI cost per query?

AI costs range from $0.0001/query for small models to $0.10+/query for frontier models like GPT-4. Use the free AI Unit Economics Benchmark (AUEB) at richardewing.io/tools/aueb to calculate your specific AI cost structure.

Product Management

2 min read

What is AI Product Management?

TL;DR

AI Product Management is a specialized discipline of PM focused on building, scaling, and maintaining products explicitly powered by machine learning, LLMs, or autonomous agents.

⚡ AI Product Management at a Glance

📂

Category: Product Management

⏱️

Read Time: 2 min

🔗

Related Terms: 3

❓

FAQs Answered: 1

✅

Checklist Items: 5

🧪

Quiz Questions: 6

📊 Key Metrics & Benchmarks

20-30%

Feature Adoption

Average percentage of features actively used

2-4 weeks

Time-to-Value

Optimal feature release to business impact

$50K-200K

Decision Cost

Cost of a wrong prioritization decision per quarter

30-50%

Zombie Features

Features with <5% monthly active usage

10x

Discovery ROI

Value of proper discovery vs. building wrong thing

40-60%

PRD Accuracy

Requirements that survive contact with users

AI Product Management is a specialized discipline of PM focused on building, scaling, and maintaining products explicitly powered by machine learning, LLMs, or autonomous agents.

Traditional Product Management focuses on deterministic behaviors: "If the user clicks this, X happens." AI Product Managers must operate probabilistically. They manage hallucination rates, precision vs recall tradeoffs, AI Unit Economics (AI COGS), non-deterministic testing, and specific prompt boundaries.

In 2025/2026, the transition from SaaS PM to AI PM demands a hard pivot toward empirical data analytics and data-pipeline architectural comprehension.

💡 Why It Matters

Treating an AI feature like a traditional software feature is guaranteed failure. AI Product Managers are responsible for the fragile bridge between raw model capability and actual user value.

🛠️ How to Apply AI Product Management

Step 1: Assess — Evaluate your organization's current relationship with AI Product Management. Where is it strong? Where are the gaps?

Step 2: Define Goals — Set specific, measurable targets for AI Product Management improvement aligned with business outcomes.

Step 3: Build Plan — Create a phased implementation plan with clear milestones and ownership.

Step 4: Execute — Implement changes incrementally. Start with high-impact, low-risk improvements.

Step 5: Iterate — Measure results, learn from outcomes, and continuously refine your approach to AI Product Management.

✅ AI Product Management Checklist

Define success metrics for AI Product ManagementMap AI Product Management impact to business outcomesCreate framework for evaluating AI Product Management decisionsBuild team alignment around AI Product Management prioritiesTrack and report AI Product Management metrics monthly

📈 AI Product Management Maturity Model

Where does your organization stand? Use this model to assess your current level and identify the next milestone.

Initial

14%

No formal AI Product Management processes. Ad-hoc and inconsistent across the organization.

Developing

29%

Basic AI Product Management practices adopted by some teams. Documentation exists but is incomplete.

Defined

43%

AI Product Management processes standardized. Training available. Metrics established but not yet optimized.

Managed

57%

AI Product Management measured with KPIs. Continuous improvement active. Cross-team consistency achieved.

Optimized

71%

AI Product Management is a strategic advantage. Automated where possible. Data-driven decision making.

Leading

86%

Organization sets industry standards for AI Product Management. Published thought leadership and benchmarks.

Transformative

100%

AI Product Management drives business model innovation. Competitive moat. External recognition and awards.

⚔️ Comparisons

AI Product Management vs.	AI Product Management Advantage	Other Approach
Ad-Hoc Approach	AI Product Management provides structure, repeatability, and measurement	Ad-hoc requires zero upfront investment
Industry Alternatives	AI Product Management is tailored to your specific organizational context	Alternatives may have larger community support
Doing Nothing	AI Product Management creates measurable, compounding improvement	Status quo requires zero effort or change management
Consultant-Led Only	AI Product Management builds internal capability that scales	Consultants bring external perspective and benchmarks
Tool-Only Solution	AI Product Management combines process, culture, and measurement	Tools provide immediate automation without culture change
One-Time Project	AI Product Management as ongoing practice delivers compounding returns	One-time projects have clear scope and end date

🔄

How It Works

Visual Framework Diagram

┌──────────────────────────────────────────────────────────┐ │ AI Product Management Framework │ ├──────────────────────────────────────────────────────────┤ │ │ │ ┌──────────┐ ┌──────────┐ ┌──────────────┐ │ │ │ Assess │───▶│ Plan │───▶│ Execute │ │ │ │ (Where?) │ │ (What?) │ │ (How?) │ │ │ └──────────┘ └──────────┘ └──────┬───────┘ │ │ │ │ │ ┌──────▼───────┐ │ │ ◀──── Iterate ◀────────────│ Measure │ │ │ │ (Results?) │ │ │ └──────────────┘ │ │ │ │ 📊 Define success metrics upfront │ │ 💰 Quantify impact in financial terms │ │ 📈 Report progress to stakeholders quarterly │ │ 🎯 Continuous improvement cycle │ └──────────────────────────────────────────────────────────┘

🚫 Common Mistakes to Avoid

Implementing AI Product Management without executive sponsorship

⚠️ Consequence: Initiatives stall when competing with feature work for resources.

✅ Fix: Secure VP+ sponsor who can protect budget and prioritize the initiative.

Treating AI Product Management as a one-time project instead of ongoing practice

⚠️ Consequence: Initial improvements erode within 2-3 quarters without sustained effort.

✅ Fix: Embed into regular rituals: quarterly reviews, team OKRs, and reporting cadence.

Not measuring AI Product Management baseline before starting

⚠️ Consequence: Cannot demonstrate improvement. ROI narrative impossible to build.

✅ Fix: Spend the first 2 weeks establishing baseline measurements before any changes.

Copying another company's AI Product Management approach without adaptation

⚠️ Consequence: Context mismatch leads to poor results and wasted effort.

✅ Fix: Use frameworks as starting points. Adapt to your team size, stage, and culture.

🏆 Best Practices

✓

Start with a 90-day pilot of AI Product Management in one team before rolling out

Impact: Validates approach, builds evidence, and creates internal champions.

✓

Measure and report AI Product Management impact in financial terms to leadership

Impact: Ensures continued investment and executive support for the initiative.

✓

Create a AI Product Management playbook documenting processes, tools, and decision frameworks

Impact: Enables consistency across teams and reduces onboarding time for new team members.

✓

Schedule quarterly AI Product Management reviews with cross-functional stakeholders

Impact: Maintains momentum, surfaces issues early, and keeps the initiative visible.

✓

Invest in training and certification for AI Product Management across the organization

Impact: Builds internal capability and reduces dependency on external consultants.

📊 Industry Benchmarks

How does your organization compare? Use these benchmarks to identify where you stand and where to invest.

Industry	Metric	Low	Median	Elite
Technology	AI Product Management Adoption	Ad-hoc	Standardized	Optimized
Financial Services	AI Product Management Maturity	Level 1-2	Level 3	Level 4-5
Healthcare	AI Product Management Compliance	Reactive	Proactive	Predictive
E-Commerce	AI Product Management ROI	<1x	2-3x	>5x

🌐

Explore the AI Product Management Ecosystem

Pillar & Spoke Navigation Matrix

⚖️ Flagship Advisory

PDI Diagnostic

Product Debt Index

Quantify the hidden economic liability of AI Product Management across your architecture using the world's leading valuation impact framework.

Deploy Tool

$10k Value

❓ Frequently Asked Questions

Do AI Product Managers need to code?

Not necessarily, but they must fluently understand data science concepts (training data, vectors, recall, embeddings) and the specific marginal costs of API token orchestration.

🧠 Test Your Knowledge: AI Product Management

Question 1 of 6

What is the first step in implementing AI Product Management?

Need Expert Help?

Richard Ewing is a Product Economist and AI Capital Auditor. He helps companies translate technical complexity into financial clarity.

Book Advisory Call →

Keep exploring

guide