Street-Smart Decision-Making
Give the customer's perception as much weight as your logic — model the narrative, not just the numbers.
- Difficulty
- Advanced
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 88%
Book-smart decision-making is everything PMs pride themselves on: data, design, technology, strategy, frameworks. Street-smart decision-making layers on top of it — giving the customer's perception equal weight to logic. A change can be provably correct on every model and prediction and still trigger a revolt because of what it looks like the moment a customer logs in. The skill is anticipating the narrative and sequencing the rollout so the perceived story isn't 'they're screwing me.'
Origin
Tal Raviv, from repeated mistakes he made and observed; he stresses the failures came from empathetic, well-meaning teams whose analyses were entirely correct.
Core principles
- 01Perception deserves as much weight as logic — this goes beyond empathy.
- 02People see the negative immediately and the positive only 30 days later; the narrative forms on the immediate view.
- 03Being right on the numbers doesn't protect you if the story feels bad.
- 04Intuition for this comes from immersion in raw customer data — support tickets, community, sales calls.
How to run it
- 1
Do the book-smart analysis first
Run the data, models, and predictions as usual. This step is necessary and Raviv's teams got it right — that's not where the failure lives.
- 2
Simulate the customer's first-glance perception
Ask: when a customer logs in tomorrow morning, what will they see and feel — before the delayed benefits materialize? Weigh that perceived narrative as heavily as the logic.
Pro tip Watch for asymmetric timing: costs/negatives visible now, benefits visible in 30 days.
- 3
Build the perception muscle from raw customer data
Spend time in the customer community, support tickets, and sales/CSM calls to learn what this persona is sensitive to (e.g. 'nickel-and-diming') and how they're quick to blame the company.
- 4
Re-sequence the rollout to fix the narrative
If the perception is bad, change how you ship — smaller pieces, different framing, staged releases — even when the underlying change is entirely above-board.
Pro tip Also apply it downstream: a 'more efficient' UX can quietly break the sales demo's aha-moment, so watch adjacent perceptions too.
Watch out Don't confuse this with capitulating to the loudest voice — see the two-week rule for separating change-resistance from real problems.
In the wild
A company had built features that were invisible on lower plans and finally wanted to reveal them so people could upgrade. The team realized customers would log in and see 'the company built everything I asked for — and all of it costs more money,' which would feel terrible. So they rolled it out in smaller pieces and different ways instead of all at once.
→ Avoided the 'they're upselling everything I requested' narrative despite nothing actually being upsold — all features already existed and everything was above-board.
Common mistakes
Trusting correct models over perceived narrative
A pricing restructure meant only to unblock a payments roadmap (not to make more money) was validated by every analysis, rolled out, and caused an internet revolt that forced a rollback. Every prediction proved correct — the miss was purely perception and narrative.
Is it for you?
Best for
PMs and leaders shipping pricing, packaging, or UX changes where customer trust and perceived fairness are at stake.
Not ideal for
Low-visibility internal or backend changes with no customer-facing narrative, where over-managing perception wastes time.
From the transcript
“it's like like giving the customers's perception just as much weight as you would to logic”
“all the analyses proved out to be to be correct all the models all the predictions like everything played out the numbers as as predicted…”
“you saw only the negative but you didn't see all the positive because the positive was you know what happened 30 days later”
“suddenly from their point of view the company built everything they' asked for and all of it requires paying more money how are they goingon…”
From the episode
Becoming a super IC: Lessons from 12 years as a PM individual contributor
Tal Raviv (Product Lead at Riverside)