RevLifter on the DTC Live podcast - How Radley grew conversion and AOV while spending less on discounts

We joined Max, head of eCommerce at Radley, on the DTC Live On Air podcast to talk about a question every retailer eventually asks: Are your promotions actually working, or are they just eating your margin?

Here are the highlights.

The numbers behind the partnership

Working together, Radley and RevLifter drove:

• 15% higher conversion rate
• 23% bigger average order value
• 8% less spent on promotional codes

None of that came from a bigger discount. It came from showing a discount to fewer people, the right people.

Stop discounting everyone

Blanket codes were the starting problem. Showing the same code to every visitor meant giving it to people who’d have bought anyway.

So Radley and RevLifter targeted a smaller group: people who actually needed a nudge to convert. The result was better conversion from that group, bigger baskets, and fewer codes used overall.

“You’ve got people on your website for a reason. There’s intent there. They don’t all need the push.”

Look past the conversion rate

Radley’s main measure isn’t conversion rate on its own. It’s revenue per user. Buying intent shifts by season and by customer, so a single conversion number stops telling the full story on its own.

Radley also tracks the smaller steps along the way:

• listing page to product page
• product page to basket
• basket to checkout complete

Those funnel metrics show where the real opportunity sits.

What actually signals intent

RevLifter now tracks more than 50 data points to work out how likely a visitor is to buy. Some are obvious, like how far someone is into checkout or whether they’ve visited before. Some are less obvious, like copying a product code to search for it elsewhere. That’s a clear sign someone is comparing prices and is close to buying.

Checkout abandonment and basket abandonment remain the most reliable signals, and the easiest place to start.

Percent off doesn’t work everywhere

Radley tested percentage discounts against flat dollar amounts in the US and found dollar-off performed better there, likely because sales tax makes the maths harder to work out at checkout. In the UK, percentage discounts still win.

The wider pattern: US shoppers lean toward cashback and card rewards over codes. UK shoppers are more used to codes, and Radley’s brand recognition there (over 90%) means less convincing is needed to close the sale.

Why blanket discounting persists

RevLifter’s research with IMRG found that around half of annual product revenue came from discounted products across its retailer panel. Discounting has been part of retail for about 150 years. It works, which is exactly why it’s hard to give up.

Some of it comes down to tech: not every retailer can issue a unique code per customer, so a blanket code is the only option available. Some of it comes down to time. eCommerce teams already juggle emails, trading, merchandising, and content, and promotions become one more job nobody has time to do well. That’s part of why RevLifter runs as a managed service rather than a self-serve platform.

Planning for Black Friday starts in May

Radley’s peak planning begins months out. Rather than widening discounts for Black Friday and Cyber Monday, they keep offers narrow and separate stock clearance from marketing promotions. They also test mechanics like a bounce-back: buy in November, get a reward to spend in December or January.

The goal is to cut through the noise with fewer, better-targeted offers rather than competing with everyone else’s blanket discounts.

Testing that actually proves something

Every campaign includes a control group, plus A/B tests on top of it, covering factors like offer type or discount level.

One change stands out. RevLifter recently moved from assigning visitors to a control group the moment they land on the site to assigning them only once they show buying signals. That single shift sharpened the results because it stopped counting people who were never going to convert either way.

“We don’t know all the answers right now. What we’re going to do is figure them out with you.”

What worked, and what didn’t

Not every test lands. RevLifter tried to nudge a gift retailer’s shoppers from buying individual products to buying an “experience” instead. It didn’t work; the two purchases were too different.

What did work: turning slow-selling stock (a lipstick shade nobody wanted, in one case) into a free gift with purchase, which drove extra sales. Prize draws, like giving one shopper their basket for free, also performed well as an alternative to blanket discounts.

Know where your codes are actually going

When RevLifter starts working with a new retailer, one of the first things they do is map every discount code in use and its source. Retailers are often surprised: affiliates, content creators, and partners are all issuing their own codes, sometimes stacking them on top of one another without anyone noticing.

You can’t manage what you don’t know exists. Getting that visibility is often the biggest single win before any targeting happens.

Keep testing simple

The biggest blocker to good CRO isn’t a lack of ideas; it’s hesitation. Radley once built a full product comparison tool, similar to the spec sheets you get with white goods, only to find almost nobody used it.

The fix: pick low-risk tests, stay focused, and keep coming back to one framework: sell more things, to more people, more often. If a test works, do more of it. If it doesn’t, drop it.

A challenge to traditional loyalty programmes

“Most loyalty programmes end with the customer earning a discount anyway. So why not skip the cards and the admin, and just give the discount to the people who deserve it?”

Large loyalty schemes work when you have the scale to fund them. For smaller retailers, a simpler approach, identifying your best customers through RFM analysis and giving them the discount directly, can get you most of the value without building the infrastructure.

What makes the partnership work

Radley treats RevLifter as an extension of their team, not a vendor to manage. Their account manager works alongside Radley’s own trading cadence, sitting in on planning and understanding what’s realistic for the brand. RevLifter offers a 90-day free trial specifically because the first stretch is about gathering data, not delivering results, and either side can walk away if the fit isn’t there.

One belief that’s gone out of date

Using discounts to clear stock that didn’t sell as forecast is a workaround, not a fix. The real problem is understanding demand before the stock gets bought. Year-on-year comparisons are also losing their usefulness. Too many changes in twelve months for last year’s numbers to mean much on their own.

“You have to plan for half your plans to be wrong.”

Contingency matters more than the plan itself. Even strong, well-run partnerships aren’t immune to the wider market. Denby Pottery, a RevLifter client until it went into administration, was performing well right up to that point. External conditions can shift the ground under a business regardless of how well a promotion strategy is running.

What’s next

Radley is pushing further up the funnel: content engagement, scroll rates, and the steps before someone even adds to the basket. RevLifter is adding more data points (past 50 and counting), building deeper CRM integrations, and expanding into more channels.

Catch the episode on the DTC Live On Air podcast, also streaming on Spotify.