The Chair on the Table: How Operational Friction Gets Mistaken for Falling Demand

When a business owner sees a number fall, the first explanation reached for is almost always the same: people want this less than they used to. It is a natural response, and it is often a misreading of the data. A meaningful share of what gets diagnosed as falling demand is actually operational friction, and it comes in two forms: created on purpose to solve someone’s own problem, or appearing by accident and going unnoticed for years. Either way, the sales figures look identical.

The distinction matters: falling demand calls for a change in offer, positioning, or price; friction, deliberate or accidental, calls for finding whoever created it and removing it. Treat one as the other, and a business can spend months solving something that was never actually broken.

The Chair on the Table

Coffee shops offer a clean example of this, identified by the behavioural economist Rory Sutherland. Sales in the last quarter-hour before closing are consistently lower than earlier in the day, and the obvious conclusion is that people simply want less coffee that close to closing time.

The real explanation is usually much narrower. Staff nearing closing time do not want a last-minute customer settling in for forty-five minutes just as they are trying to clean the machine and cash up. The fix, passed between shop workers rather than written down anywhere, needs no sign and no change to opening hours: stack one chair upside down on a table, in plain view of the door, or lean a mop against the counter.

A customer walking up reads the scene in an instant: the staff are winding down for the night, and walking in now would mean holding them up, or nursing a coffee half-watched while someone waits to mop around them. Rather than settle in and feel rushed, most people simply turn away. Nothing about the shop has actually changed.

The sales dip is real. The decline in demand is not. The chair, or the mop, would never appear in the data, and whoever reads the report has no way of knowing it was engineered, not decided by the market.

When the Numbers Look Like Falling Demand, But Aren’t

Sales data cannot tell a customer who decided against buying from one who never got the chance to. Both produce an identical number. This is precisely why operational friction is so easy to miss: it hides behind a metric that looks exactly like waning interest, and most reporting stops there.

The businesses that catch it are the ones that keep asking what happened immediately before the number moved, rather than treating it as the final word.

The Twelve-Million-Dollar Field

Expedia found a version of the same problem buried inside its own booking form, though nobody had built this one on purpose. The company noticed something odd: plenty of customers got all the way through the process, entered their details, and clicked “Buy Now”, about as clear a sign of intent to buy as a business will ever see, and the booking still never went through. That kind of gap is exactly the number a business might write off as last-minute hesitation, or a price that suddenly felt too high.

Rather than accept that, the team looked at what was actually happening on the page itself. Sitting underneath the customer’s name was an optional field labelled “Company”. A significant number of people mistook it for a request for their bank’s name, filled it in accordingly, and then, following the same logic, entered their bank’s address in the field meant for their own home address. When the card issuer checked that address against the one it held on file, the mismatch caused the payment to fail, through no fault of the customer’s own.

Deleting the single field produced an immediate, sustained increase of twelve million US dollars in annual profit. No pricing changed, no product changed, and the demand had been there all along. The field had sat there for years, nobody meant for it to cost anything, which is exactly why nobody had questioned it. Joe Megibow, Expedia’s vice-president of analytics and optimisation at the time, said the team found fifty or sixty similar problems just by paying attention to where customers dropped out.

Diagnosing Operational Friction Before Blaming Demand

Both cases point to the same discipline: check for friction, chosen or accidental, before concluding that appetite has genuinely fallen.

The Takeaway

Falling numbers invite an obvious story: customers want this less than they used to. Sometimes that is correct. Just as often, a business has quietly made itself harder to buy from, on purpose or by accident, and the gap gets misread as a verdict on demand rather than a signal worth investigating.

Getting this distinction right changes where a business spends its energy. Instead of discounting a product customers never stopped wanting, the fix can be as narrow and as findable as a single form field, or a single chair left the right way up.