I have loyalty cards for three supermarkets and cycle between them for offers — has this saved anyone real money?

The part that got me was the basket effect. When I shopped for offers I bought things because they were discounted rather than because I needed them, and my monthly grocery spend went up in the months I was most diligent about the points. I only spotted it because I had two years of data and the correlation was ugly. Saving 30 percent on something you did not need is a 70 percent loss.

I tracked mine for a year because I had the same suspicion. The base earn rate on points came out under one percent of what I spent, which is worth having and is not worth planning your week around. The member-only prices were a different story and were worth several times the points — but only on a handful of items, and only when I would have bought them anyway. Once I subtracted fuel for the extra trips, the whole multi-shop routine was roughly break-even and cost me two hours a week.

One shop, one list, no apps, and buy the cheap version of the boring items. Beats every loyalty scheme I have ever tested and takes no time at all.

Partial disagreement with the dismissive answers. Blanket points chasing is close to worthless, agreed. But the targeted personalised offers some schemes push at you are genuinely valuable if you only ever activate the ones on your normal list and ignore everything else. I do about three minutes a week on that and it is meaningfully more than one percent. The trick is that it takes three minutes, not two hours, and involves no extra driving.

Redemption is where these quietly leak. Points that expire, points that are worth more against some categories than others, points sitting in a scheme for a shop you no longer use. Spend them regularly and small rather than saving them for something impressive, because the balance you are proud of is the balance most likely to be devalued.

Work out the value of your points balance in pounds, then divide it by the hours you spent earning it. Mine came out at an hourly rate I would not get out of bed for, and that ended the debate in about four minutes.