I once watched a company spend five years building a world-class e-commerce busi
By Faraz Siddiqui · October 7, 2026 · Curated by George's Blog
I once watched a company spend five years building a world-class e-commerce business, and then dismantle it in a single re-org that everyone involved called progress.
The logic was clean, the way dangerous logic usually is.
E-commerce had grown up as its own scrappy unit, with its own P&L, its own odd metrics, and its own people who spoke a language the rest of the business found tiresome.
Now that it was big and it mattered, the obvious move was to fold it into retail.
Simplify. Align. One commercial team, one way of working.
The slide practically applauded itself.
Here is what the slide did not show.
The people who had built that business did not think like the rest of the business. They thought in lifetime value, shopper cohorts, and repeat purchase. They tracked the slow compounding of a customer who comes back on their own.
The retail leaders they now reported to thought in weekly sell-in and monthly sell-out, Nielsen market share, promotional calendars, and availability on the shelf.
Both are real skills. They are not the same skill.
And one of them had just been put in charge of the other.
So the questions changed. "How do we build a customer who returns without a discount?" became "What's the promo that hits this week's target?"
The e-commerce team's patient metrics were labeled overcomplicated and retired.
Within two quarters, the best of those people understood exactly what had happened, and left. People who think in years do not stay long in a job measured in weekends.
The revenue held for a while. It usually does.
That is the part that has stayed with me.
Nothing broke on the day. The re-org simply removed the hard-won muscle that had created the growth and replaced it with one built for a different sport.
If I were asked to design that re-org today, I would merge the org chart first and the metrics last. The team inheriting a capability should have to learn how it measures success before they are allowed to change it.
And the people who built it should be kept close to the decisions for a year, not two layers below them.
Capability is a slow thing to lose. The damage hides inside numbers that still look fine, right up until the growth that always felt automatic stops arriving, and nobody left in the building knows how to go and get it.
Companies love to consolidate a capability the moment it finally starts working. Sometimes they are capturing a synergy.
Sometimes they are losing the people who knew why it worked.
The org chart will never admit it.