Chapter Fifteen

The “Should We
Use AI?” Framework

Start with the user, not the model. This is the checklist you tape to your monitor the next time someone says “we should AI this.” It has killed more shiny-but-pointless features than any other tool PMs adopted in 2025–2026.

📖 ~12 min readPages 102–107
scroll
Maya's numbers
17%
adoption. Time-to-inbox-zero went up 11%.

Maya sat in the war room at 9:47 p.m., staring at usage data for her AI email triage feature. Two months ago: “Your inbox, now with 40% less noise.” Users opened it once, played with confidence sliders, then went back to manual sorting. Adoption: 17%. Time-to-inbox-zero went up 11% — users spent extra minutes second-guessing the AI.

Maya didn't fail at building AI. She failed at asking the right question. The job wasn't “triage with AI.” It was “quickly separate urgent from ignorable without missing anything that burns me.” Simple saved searches, shortcuts, and smarter snooze would have crushed 80% of the friction with zero hallucinations.

The 80% ruleSaved searches + shortcuts + snooze = 80% friction removed. Zero hallucinations.

Start with the User, Not the Model

Friction Audit

Write the JTBD in one sentence, no tech words. Then list every painful step:

Key insightThe differentiator is judgment about when NOT to use AI.

Cognitive: “Which of these 47 threads is actually urgent?”

Time: “Reading every subject line at 6:12 a.m.”

Error: “Missing the one email from legal that costs $40K.”

Emotional: “That sinking feeling when the badge hits 200.”

Only after this: Does AI actually remove a friction, or replace it with “did the AI get it right?”

OPEN SCAN TRIAGE ACT Badge anxiety 47 subject lines Miss legal email ✓ RULES / UX CAN FIX ❓ AI TEMPTATION ZONE Most friction lives on the left. AI is rarely the first answer.
Figure 15.1 — Friction-First Audit. Map the workflow. Red = pain. Green = what rules solve. Only the remainder is the AI conversation.

User-First AI Scorecard

Score every AI idea 1–10. Total <32? No AI. 32–38? Rules first. 39+? Talk models.

The 60% killThree teams scored their AI backlogs. 60% of ideas died in one afternoon. Survivors shipped faster.
1Can rules / UX solve 70%+?Rules win on determinism, speed, explainability. Maya's case: priority labels + keywords = zero model needed.__ /10
2Does AI reduce real effort, or add novelty?$5 test: Would users pay $5/mo if 100% accurate? If “maybe for the wow” — kill it.__ /10
3What breaks when it's wrong?Cost-of-wrong = frequency × severity. If >$10K/mo at scale, human-in-loop or walk.__ /10
4Do we control data + tolerate non-determinism?If the job requires perfect consistency (finance, legal), AI needs heavy scaffolding.__ /10
5Can we define success WITHOUT accuracy?Adoption, trust, net time saved, decisions improved — the only metrics that matter.__ /10
<32: NO AI
32–38: rules first
39+: models
Figure 15.2 — Scorecard. Three teams adopted this and killed 60% of their AI backlog in one afternoon.
The $5 testWould users pay $5/mo if 100% accurate? “Maybe for the wow” = novelty, not value.
Chapter 15

Unlock the full chapter

The first two chapters are free. Chapters 3 through 18 unlock with a one-time purchase on the same account.

$18.99one-time purchase via PayPal

Already purchased? Sign in with the same account you used at checkout.