The mechanism of ordering a product can matter more than the product itself — small psychological bottlenecks can doom a business, and fixing them (like offering a postal option alongside phone ordering) can double response rates in ways that economic logic cannot explain.

Rory describes a randomized test for British Telecom offering call-diverting services: postal-only got 5%, phone-only got 2%, but offering the choice of either produced 7% — the sum exceeded the parts. This taught him that brilliant businesses fail because they never unblock a psychological bottleneck, and that direct response advertising discovered behavioral economics decades before the academics. ✦ AI generated

Rory Sutherland · My First Million · 2026-07-27 · original ↗

I'll give you an example. This is it's something like 1991, 1992, and I'm working on direct response advertising for the phone company. And we simply um were offering people the chance to pay a couple of pounds a month extra to receive what what we call star services I think in the United States network services which is you could have call diversion you could have call waiting on your phone and you could have something else wake up call or something like that and and if you've paid a certain amount monthly these facilities were added to your home phone line and we wrote people letters and it said either send back the pre-lasered coupon below in the in the postage paid envelope or call this number and quote this code. We gave them a choice of post or phone. And we had a slightly bonkers client who didn't want to offer the postal response. I don't know why. I have no idea why to this day. Okay. And so we said, well, let's test it. um let let's not just get rid of a coupon response because that's dangerous because you know if you know if you don't test this rigorously you could end up effectively damaging your business quite significantly on a whim. And so we tested three kinds of letters, 50,000 people each, randomized, completely randomized selection as with the newspapers. And one lot got postal only, one lot got phone only, and one lot as before got the choice of phone or post. Post only 5% response rate. Phone only 2% response rate. When you offer people a choice, it was just it was 7%. not quite but almost nearly I think the sum total of the of the two independent response rates that that may not surprise strike you as all that weird but to an economist this would be completely baffling because it would suggest that the more important factor affecting whether you bought the product was not what the product was or how much it cost but how you were able to actually order it. And from that moment on I remember thinking okay this is extraordinary because every now and then there are almost certainly brilliant businesses which are failing because they've failed to unblock some sort of psychological bottleneck. At this case, the psychological bottleneck might have been that, you know, all you know, people just hated using the phone to order something. And you could have rationally spent millions and millions and millions of pounds trying to improve the product or to spend ages demonstrating what a great product it was. And yet, if you fail at one psychological hurdle, you can doom yourself.

verbatim transcript · starts at 36:49