When Figma strictly enforced AI credit limits in early 2026, user feedback was swift: current allocations felt tight, prompt consumption felt unpredictable, and usage limits lacked transparency. Yet, market signals reveal that power users aren’t rejecting paid AI outright – they are demanding a fair, predictable, and controllable monetization structure. To uncover what users will actually pay for, EPIC conducted a choice-based conjoint study of 225 current Figma Make users, evaluating how they trade off monthly credit allocations, rollover rules, efficiency tools, spending controls, and price.

The findings challenge conventional SaaS packaging logic. While credit allocation is the largest single driver of choice (28.4% relative importance), spending control and credit efficiency features combined match its impact at 28.9%. What earns a paid upgrade isn’t simply a larger bucket of credits, but the tools to see, manage, and stretch those credits in real time. Features like automatic usage alerts, per-prompt cost estimates, draft modes, and smart routing offset price resistance, dramatically reducing plan rejection and turning community friction into upgrade revenue.

Crucially, the study demonstrates why credit-heavy tiered portfolios can fail on AI-cost-adjusted margins. While multi-package lineups generate high gross subscription revenue, large credit allowances carry substantial unit inference costs. When accounting for estimated AI usage costs, an optimized single plan ($30/month for 3,000 credits paired with complete control and efficiency features) yields the highest net returns. Read the full whitepaper to explore the complete choice-based dataset, preference curves, and unit-economic models for AI product monetization.