Ritson says there are ten reasons never to discount. He’s right, and he’s wrong.

Mark Ritson has written a very good article about a problem most retailers don’t have.

His piece in The Drum lists 10 reasons never to cut prices. It’s sharp, funny and mostly correct. Blanket discounting does destroy margin, and it does train shoppers to wait. It does turn your RRP into fiction.

We agree with more of it than you might expect, considering discounts are our thing.

But the article has a blind spot, and it’s a big one. Ritson builds his case on P&G. A brand owner running a 24.3% core operating margin in FY2025, holding no stock on its shelves, with the luxury of waiting years to win back share.

Most retailers have none of those things.

The retail sector runs a net margin of around 3-6%. Retailers own the stock and pay for the warehouse where it sits. And every week it doesn’t sell, it’s worth less.

That’s not a rounding error between two business models - it’s a completely different set of rules.

The maths cuts both ways

Ritson opens with the arithmetic:

Cut 20% off a typical consumer margin, and you need to roughly double volume to stand still.

True, but incomplete.

Because that calculation assumes the alternative is selling the same unit at full price. For a lot of retail stock, it isn’t - the alternative is to sell it at 50% off in eight weeks or not sell it at all.

Research by Coresight and Celect put the cost of markdowns to US non-grocery retailers at around $300bn in a single year. About 12% of all US non-grocery retail sales. The same study found the average full-price sell-through rate is 60%. Four out of every ten units need a discount to move.

You can see it play out at the brand level. Analysis of eight national US chains found they discounted 44% of their Nike sneakers on average in 2024, up from 19.4% two years earlier.

Nobody planned that. It’s what happens when supply and demand no longer match, which is most of the time. Forecasting isn’t easy.

So the real question isn’t “full price or discount?” It’s “which discount, to whom, and when?”

A 15% offer to a hesitant shopper today is cheaper than a 50% clearance in October. Getting that right is a trading skill. Refusing to engage with it isn’t discipline, it’s just a slower write-off.

He’s right that most promotion is wasted

Here’s where we’d hand Ritson the microphone. If he promises to behave.

His third and fourth points were that promotional spikes come mostly from people who’d have bought anyway, and that you’re borrowing from next quarter, and are the strongest in the piece.

The data backs him. Research from Inmar Intelligence is widely cited for the finding that around 72% of promotional spending generates no measurable incremental lift. Practitioner analyses put cannibalisation on broad public coupon codes at 20-60%. That’s the share of discounted orders that would have happened at full price anyway.

An enormous amount of margin was handed to people who already had their cards out.

But look at what that number is actually describing. It’s not proof that promotions don’t work. It’s proof that untargeted promotions don’t work.

Targeted offers behave differently. Cannibalisation in areas like abandoned-cart recovery or new-customer-only codes runs at 10-25%.

There’s a good published example. A large US home decor retailer ran an A/B test across 20 million customers, comparing two ways of deciding who gets which discount. Same promotional budget on both sides. The better-targeted approach delivered 4.5% higher revenue and a 5.6% increase in contribution margin.

Not one extra penny of discount, just aimed at the right targets and the right time.

Ritson’s answer to a 72% waste rate is to stop spending. Ours is to stop wasting.

You’re already discounting - you just call it something else

This is the part of the argument that doesn’t hold up under scrutiny against a real retail P&L.

• Free shipping over £50? That’s a discount with a condition attached. Thresholds are reported to lift average order value by 12-24%, with 58% of shoppers adding items to qualify.
• Loyalty pricing? The CMA found UK supermarkets took over £5bn in loyalty-priced revenue between November 2023 and January 2024. Around 22% of total grocery revenue for the chains that run it. Reviewing about 50,000 loyalty-priced products, the CMA found 92% offered genuine savings of 17-25%.
• Bundles, multibuys, welcome offers, student codes, gift-with-purchase, points that convert to money off?

All discounts.

The choice was never between discounting and not discounting. It’s between discounts that ask something of the shopper and discounts that don’t.

A conditional offer changes behaviour. Spend more. Buy the slower line. Come back within 30 days. Join the programme. Take the size we’re overstocked on.

An unconditional offer just moves money from your margin to a shopper who’d already decided.

Ritson attacks the second and assumes it’s the whole category.

Promotion is a merchandising tool, not just a pricing one

The article treats a discount as a single lever with one setting: on or off.

Retail doesn’t work like that.

The same 20% means completely different things depending on what it’s attached to. Put it on your hero line, and you’ve burned margin on guaranteed sales. Put it on the size 14s you’ve got 900 of, shown only to people browsing that category, and you’ve solved a stock problem without touching your headline price.

Same discount. Opposite outcome.

This is where his sixth point, that promotions annoy your loyal base, gets interesting. He’s right that they do. Worth noting the CMA found 40% of shoppers don’t trust loyalty prices are a genuine saving, even when the evidence says they are. Shoppers are watching, and they are sceptical.

But the fix isn’t abstinence. It’s exclusion.

Don’t show the code to the customer who’s already at checkout with full-price intent. Show it to the one who’s bounced twice this week.

You can’t do that with a sitewide banner. You can do it with rules.

Black Friday isn’t going anywhere

Ritson uses Black Friday as his exhibit for the prosecution. Margin sacrificed, stockpiling encouraged, a sales lull to follow.

Fair, and also: Black Friday 2025 drove a record $11.8bn in US online spend, up 8.7% year on year, with Salesforce putting global sales at $79bn.

The average discount rate was 28%. Exactly the same as the year before.

Which tells you something. The winners weren’t the ones cutting deepest. They were the ones cutting most precisely, at a moment when shopper intent was already high, and acquisition costs were low.

A retailer who sits out Black Friday on principle doesn’t preserve their brand. They just hand the day to a competitor and end up paying more for the same customer in January.

Where we’d actually agree with him

Plenty, as it turns out:

• Blanket, always-on discounting is a slow death. Yes.
• Reference prices, once lowered, are hard to raise. Yes.
• Discounting because a competitor did is not a strategy. Yes.
• Value is what the shopper gets divided by what they pay, and improving the numerator is the better long game. Completely yes.

Tide’s upgrade is a good story. Better product, same price, brand back in growth.

But P&G sells detergent through other people’s shops. The retailer stocking that Tide still has to clear last season’s range, hit a weekly sales number, and compete with the shop next door that just went 25% off.

“Build the brand for years” is excellent advice. It’s also not available to a merchandiser on a Tuesday with three weeks of cover and a sell-through problem.

The rule we’d write instead

Ritson’s rule is: never discount.

Ours is shorter and harder to follow.

Never give away margin you didn’t need to give away.

That means every offer earns its place:

• Does this shopper need an incentive to convert, or were they converting anyway?
• Is the offer solving a real problem, e.g., stock, cover, acquisition, basket size, or just filling a slot in the calendar?
• Is there a condition attached that changes behaviour?
• Can you turn it off tomorrow without your sales falling through the floor?

If the answer to that last one is no, Ritson has already won the argument, and you should read his piece again. If it’s yes, you’re not discounting. You’re trading.