Most people, including a lot of senior leaders, think pricing is a number. They want to know what the number is, why it changed, and who approved it. After several years building international pricing at Walmart, I've come to a different view: pricing is a product. The number you see on a shelf — physical or digital — is the output of dozens of upstream decisions, models, and trade-offs. The product is the system that produces it.
The competitive signals you trust (and how stale they are)
The guardrails that stop a bad price from going live
The override workflow for category managers
The post-launch readouts that close the loop The shelf price is the very last mile. Most of the value sits upstream. Teams that fight about the last mile are usually losing the upstream battle.
Silent staleness. A pricing model that "still works" on the dashboard but is quietly using a competitor feed that hasn't refreshed in a week. Symptoms appear in margin, not in the system.
Invisible overrides. A category manager overrides 10% of prices, and nobody centrally knows it's happening. The model gets blamed for the leftover 90%.
The unfalsifiable launch. A new pricing capability ships without a clean experiment design, so when results are mixed nobody can tell whether it worked. The fix for all three is the same: build the boring observability layer before the exciting model. The model will get attention from leadership. The observability layer will save your job.
A "small" change to your model in one market can quietly break another market that depended on a side-effect of the original behaviour.
Local merchant intuition is a real signal. PMs who treat local teams as a feedback channel — not a request channel — get better launches.
The metric that matters in market A is sometimes not the metric that matters in market B. Don't ship one definition of success.
A short note for PMs entering pricing
IfyouaresteppingintoapricingPMrole:
Spend a week reading historical incidents before touching the roadmap. They are the cheapest education you'll get.
Make friends with finance early. They will see your problems before your dashboards do.
Don't try to "win" the override debate. The override debate is permanent. The job is to make overrides observable, not to abolish them. Pricing rewards people who treat it as a system. The number is just the receipt the system prints out.
