I copied my neighbour’s idea and forgot the ten that closed

1The queue

Composite example, not a specific client

A new concept on one of Riyadh’s boulevards. Call it what you like: burgers, specialty cookies, “Najdi fusion”, whatever this year’s wave happens to be. It has a queue every night, and its videos are on the partners’ phones.

At the Sunday meeting someone says, “Why don’t we do one? We have the staff, we have the sites.” Six months later the group opens its version: the same fit-out mood, the same price point, a similar name.

The question nobody asked at that meeting: how many shops opened with the same idea on the nearby streets… and closed?

2The planes that came back

The story as it’s usually told: in the Second World War, analysts looked at returning bombers and saw where the bullet holes clustered. The natural proposal was to add armour where the holes were.

The statistician Abraham Wald’s answer: put the armour where there are no holes, because the planes hit there never came back to be counted.

It’s more than a tale. In 1943 Wald wrote technical memoranda for the Statistical Research Group with a method for estimating a plane’s vulnerability from the damage on survivors, allowing for those lost; Marc Mangel and Francisco Samaniego explained the work in 1984.¹ The famous red-dotted plane you see everywhere is a modern illustration, not from Wald’s papers.

The boulevard in Riyadh is the airfield. You’re counting holes on the planes that came back.

3Why risk looks like a recipe for success

In 2003 Jerker Denrell described the trap precisely: firms learn from those they can see, and those they can see are the survivors. In a sample of survivors only, risky practices can look like the cause of success even when they aren’t, because the firms that tried the same practice and fell have dropped out of the sample.²

Think of it this way. A bold concept produces a few big winners and many quick losers. The losers vanish from the street within months, so you walk down the boulevard, see only the winner, and conclude that boldness is the recipe.

And how many actually vanish? A correction is needed here. The popular line “90% of restaurants fail in their first year” has no source. H. G. Parsa’s team traced it to a 2003 American Express advertisement; when the company was asked for its data, it said in writing that it couldn’t provide any to support the claim.³

In their data, from more than 2,400 restaurant licences in Columbus, Ohio, between 1996 and 1999, about a quarter failed in the first year, and about 60% within three years. Their “failure” includes a change of ownership, not only closure.

The truth is kinder than the myth, and still harsh. But those are American figures from the 1990s that I won’t carry to Riyadh, and there’s no published Saudi figure on restaurant closures or on how long concepts survive. The point isn’t the number; it’s that you don’t know the survival rate for your concept, and the boulevard won’t tell you.

4The autopsy: fourteen months

The same composite example; the details are illustrative.

Months 1 to 3: the launch. A curiosity rush, videos, a small queue. Sales above plan. In the group chat: “Told you so.”

Month 4: the dip. The curiosity is spent; whoever came to try it has tried it. Fewer come back than the plan assumed, because the plan was built on the neighbour’s queue, not on this branch’s customers.

Months 5 to 9: the discounts. A promo code on the apps, then a “two for one”, then paid visibility. Sales hold; contribution falls.

Months 10 to 13: “evolving the concept”. A new item, a tweak to the name, a second campaign. The problem was never the item.

Month 14: the closure. And a “For rent” sign on the shopfront.

What nobody counted at any point: on the surrounding streets, before we even started, shopfronts where copies of the same idea had closed had been re-let to other businesses. They were right in front of us. But a re-let unit doesn’t look like the grave of an idea; it looks like a new shop.

None of this is a mark against those who closed. They tested the idea with their own money, and we could have learned from them for free, had we looked.

5Where the money leaks

The fit-out on a borrowed idea is sunk on opening day. None of it comes back if you close, and money already spent is no reason to stay, as last season’s finale argued.⁴

The bigger leak isn’t in the copied branch. It’s the capital and attention pulled away from your proven formats. In the doctrine, an unproven unit earns a pilot; a proven unit earns a budget.⁵ A copied concept is unproven in your hands, however long the queue is at your neighbour’s.

6The Sunday-morning question

One question, one metric, one rule

The rule isn’t mine: the definition of a “proven format” and the experimentation envelope both come from the doctrine.⁶ The 10% is its starting point; your group sets its own.

8Credit, and the next episode

The planes are Abraham Wald’s, the survivors’ trap is Jerker Denrell’s, and the 90% myth was taken apart by Parsa’s team. The boulevard, the queue and the re-let shopfront are Riyadh’s.

And sometimes the failed copy doesn’t come from your neighbour… but from your own head office, which keeps growing even when it produces nothing. That’s the next episode: “Your head office is growing faster than your branches.”