Tokenomics for the Deeply Skeptical | AI Spend, ROI, and What Engineering Leaders Should Measure

AI budgets are under scrutiny earlier than most engineering leaders expect. Token spend is up across engineering orgs — and the conversation about what it's buying is coming faster than most leaders have prepared for. In this session, Uplevel's Amy Carrillo Cotten and MC Johansen are joined by veteran engineering executive Ed Quick to break down what token management actually measures, where AI investment produces returns, and what the orgs getting real results did differently from the start. What you'll learn: — Why token spend became the default AI metric — and what it's actually telling you — How to define AI ROI for your specific org before someone else defines it for you — What good AI governance looks like in practice, at both enterprise and mid-market scale — How to make strategic AI bets when the budget has real limits — What to say to your CFO before that meeting is on the calendar Ed walks through a program that delivered 300% productivity improvement year over year with fewer engineers — including what made it work, how buy-in spread organically, and where the token spend actually went. Amy shares Uplevel data on where AI investment plateaus across engineering orgs and what the teams breaking through have in common. Timestamps: 00:00 Introductions 02:00 What the agentic SDLC actually requires 04:00 Uplevel data: what AI is and isn't moving 08:00 Discussion begins 08:05 Why token spend became the default metric 12:00 How to define AI ROI — and who should be at the table 16:00 The factory floor mindset and why it derails the AI ROI conversation 20:00 Ed's three-skill sprint: 300% productivity improvement year over year 26:00 Framing the experiment around value, not AI usage 28:00 Strategic choices with a constrained token budget 30:00 Token management vs. cloud cost governance — where the analogy holds 34:00 What to say to your CFO about AI spend 37:00 Engineering leadership is learning in real time too 38:00 Q&A: governing AI spend across teams at different maturity levels 44:00 Q&A: how to make the case for measuring outcomes when leadership wants adoption numbers 47:00 Next steps Learn more about how Uplevel works with engineering organizations: uplevelteam.com Free AI maturity diagnostic — no tool integration required: uplevelteam.com/stackup