Graybridge AI · AI product · Senior · 45 to 60 minutes

Pricing AI usage as workloads vary

Graybridge AI is a hypothetical company that sells an AI workspace for teams to run chat, document tasks, and autonomous agent sessions. Answer this product management case and get it graded line by line against a weighted rubric.

Pricing strategy
Monetization
AI product judgment
Executive communication
Risk management

The situation

You are the senior PM for monetization at Graybridge AI. The CEO and CFO have asked you to recommend how the company should price product usage before next quarter's planning cycle. Today, customers buy subscriptions that include a set amount of request-based usage, but product behavior is changing. Some workflows are still simple, short tasks, while others are long agent sessions that take many steps and use much more model capacity. Support hears complaints from some customers that basic work can feel overpriced, while finance sees that heavier autonomous usage can put pressure on margins. Sales is worried that any change that is harder to understand will hurt self-serve conversion and make enterprise renewals harder. Assume the company must keep serving both casual users and teams that rely on long-running automation. Your job is to recommend whether to keep the current model or move to a different pricing structure, and how to roll it out. The tension is balancing cost alignment, customer trust, spending predictability, and the risk that pricing changes alter product adoption and usage patterns.

What to produce

A 1 page strategy memo, 700 to 900 words.

Your answer must

State a clear recommendation on whether to keep flat request pricing, move to a usage-weighted model, or use a hybrid, and explain why.
Analyze at least three customer usage patterns, including short tasks and long autonomous sessions, and describe the likely impact on each.
Explain how your pricing approach maps to underlying cost drivers while remaining understandable for both self-serve and enterprise buyers.
Provide a rollout plan covering new customers, existing customers, contract renewals, and customer communication.
Define 4 to 6 success metrics and guardrails, plus 2 major risks and mitigations.

Constraints

Do not assume the company can materially reduce model costs or change core product behavior in the next quarter.
Existing enterprise contracts cannot be repriced before renewal.
The pricing model must work for both self-serve and sales-led plans.

Format

Use headings: recommendation, rationale, customer impact, rollout, metrics, risks.

How Graybridge AI actually played it

Explores credit-based pricing for AI products, including a June 2026 change from flat request units to usage-weighted credits at one cent each. It explains how credits act as an intermediate currency that changes how customers judge value and fairness.

Credits change how users budget and judge fairness
Tie credits to cost drivers but keep them understandable
A long autonomous session can consume many credits

Built from The Double Conversion Problem in Credit-Based AI Pricing Models by The Pricing Conundrum.