Discounts buy
the wrong
customers.

Why is discounting the wrong way to grow a DTC brand?

Because it brings in the shopper who came for the number. A discount-led acquisition strategy fills your file with deal hunters, and they leave when someone else has a bigger number. New brands also can't win that race, because the incumbents can go lower for longer than you can.

I don't think acquiring deal shoppers is sustainable in DTC, and it attracts the wrong audience. The best brands are built by leading with value and storytelling. That story is what lets them charge full price, and full price is what pays for the next customer.

Almost nobody builds the offer that way. In our capture of 1,494 DTC welcome offers, only 3.3% said anything about the second order, the order that decides whether the first discount paid off. Everyone is paying for order one and hoping order two shows up.

The alternative is an incentive that costs less than a straight discount and points at the next purchase. Each one below moves a specific letter in RFM, the recency, frequency and monetary view of your customers that finance people use to judge a consumer brand.

What does a discount actually cost a DTC brand?

Three things: margin on the order, the type of customer it selects, and the habit it teaches. Take an illustrative brand with an $80 average order and 60% gross margin. That order earns $48 of gross profit. A 20% code takes $16 off the price, so it gives away a third of the profit to win the sale.

The margin cost is the one everyone sees. Selection lasts longer. A sitewide code attracts people who respond to codes, and those customers are the least likely to come back at full price. I think a sale is only value if you'd have bought at full price anyway. For a lot of the people a code brings in, the code was the reason they bought, so it's also what they'll wait for next time. The cohort evidence is in the Black Friday version of this argument.

The third cost is the habit. Once your list learns that a code turns up every few weeks, full-price buyers start waiting for it, and more of each discount goes to customers who would have bought anyway. You see it later as falling margin per order.

Price incentives still have a place. Build them to change what the customer does next, like placing a second order. For the math on how a discount trades conversion against order value, see AOV vs conversion rate.

What can you offer instead of a discount?

Offers that pay out later, reward a bigger basket, or give access instead of a lower price. The table maps five of them to the RFM letter each one moves. R is recency, F is frequency, M is monetary value.

1. Store credit with an expiry date (the Kohl's Cash model)

Kohl's runs the best-known version. During its earn periods, shoppers get a $10 Kohl's Cash coupon for every $50 they spend. The credit only works during the redemption dates printed on it, not before and not after. Shoppers leave the first order already holding a reason to come back, with a deadline attached.

The expiry date is the point. It pulls the second order closer and shortens the gap between purchases, which is recency and frequency together. Credit is also cheaper than it looks, because you give it at retail value, it costs you the product cost, and some of it goes unspent. Watch whether customers only buy inside your credit windows. If they do, you've built a sale calendar with extra steps.

2. A reward on the second order, not a discount on the first

HexClad's popup, as Ethan Denney of ConvertFlow described it in a December 2025 LinkedIn post, offered $75 off a second purchase where most brands offer 10% off the first. The first order keeps its full margin on a $999 flagship set, the product doesn't look cheap, and the reward only pays out to someone who comes back. That is what they ran then. HexClad's VP of Global Creative replied on that post that they'd "changed A LOT this year", so treat it as an example, not their current offer.

This moves frequency. Claiming a future reward also tells you who plans to buy again, which a first-order code never does, and those are the people worth more follow-up across email and SMS. It fits a considered first purchase with a high price, like cookware or furniture. If your products last for years, make sure there's something else in the range to spend the reward on.

3. A limited window to buy back seasonal or retired products

Le Labo's City Exclusive fragrances are usually sold only in the city they belong to. Once a year they go on sale worldwide for a short window: samples from August, full sizes in September. Nothing is marked down: the reward is access to something people usually can't buy.

Most DTC brands don't have regional products, but most have a back catalogue: a retired colourway, last winter's scent, a limited run that sold out. Open a short window for existing customers to buy those back at full price. It moves frequency and monetary value, and it gives your best customers a reason to open your emails. Only promise what you can make or still have in stock.

4. A reward that only starts above your current AOV

Set the threshold above where customers already spend. If your average order is $80, the reward kicks in at $100, roughly 25% higher. Free shipping, a gift or a credit all work. The customer who was going to spend $80 adds one more thing, and the extra basket pays for the reward.

This moves monetary value. Set the threshold from your real AOV, not a guess: a threshold below where people already spend gives the reward to orders you were getting anyway. Then check the margin at the threshold: the reward should cost less than the gross profit on the extra spend it brings in. Contribution margin per order is the number to run it against.

5. Gift with purchase or early access

A rotating gift with purchase, or early access to new drops for existing customers, gives value without lowering the price. Both move frequency when they rotate, because there's a new reason to come back each time. I cover both, with brand examples, in the Black Friday piece.

Do you need a loyalty program to get repeat orders?

Usually not to start. Every brand I've ever met with who brings up loyalty programs wants to have one that's sticky like Starbucks, but this falls apart when I ask them if they have the frequency and LTV of a Starbucks customer. Most don't, and a points program built for daily coffee buyers won't change that.

A points program needs frequency to work. Points only feel valuable when customers earn them often enough to see a balance build. If your customers buy two or three times a year, they forget the balance between orders, and you've paid for software, design and a liability on your books for a habit that never forms.

One of the simple offers above often does the same job. Store credit with an expiry is a one-step loyalty program. A second-order reward is a loyalty program for the order that matters most. Start there, measure the second-order rate, and build the program later if the data says customers want more reasons to come back.

If you already run one, test it properly. Our repeat purchase benchmarks explain why loyalty member numbers overstate the lift, and how a holdout group gives you a clean read.

What is RFM, and why does it matter in a raise or an exit?

RFM scores each customer on three things: how recently they bought, how often they buy, and how much they're worth. Finance people look at it on consumer brands because it shows how connected customers are to the product. That's the best guide to whether revenue repeats or has to be bought again every month.

• Recency: when the customer last bought. A customer who ordered last month is more likely to order again than one who ordered last year.
• Frequency: how often they buy. Two orders is the big step; most customers never make one.
• Monetary: what they're worth over time, close to LTV. Measure it as gross profit, not revenue, as LTV vs lifetime revenue explains.

In my experience, RFM can materially change a deal: a fundraise, an exit or a partnership. Take two brands with the same revenue. The one whose customers come back on their own is worth more, because the other one's revenue stops when the ad spend does. Buyers usually test this in commercial diligence rather than in the quality of earnings review, and a file full of discount-acquired customers shows up there.

That is why the offer matters past this quarter. Every discount-acquired customer who never returns drags recency and frequency down for the whole file. Each second-order reward that gets claimed pushes them up. If you plan to raise or sell in the next two years, your offer strategy is already writing the cohort tables a buyer will read.

How do you compare discount-acquired and full-price customers?

Split customers by whether their first order used a discount code, then compare RFM for the two groups over the same window. If the full-price group comes back sooner, buys more often and is worth more, your discount is buying the wrong customers, and you now have a number to show for it.

1. Export customers with their first order date and whether that first order used a discount code. Shopify's customer and order reports have both.
2. Keep only customers acquired 12 or more months ago, so every customer has had time to come back. Use a longer window if your product lasts longer.
3. For each group, work out recency as days since the last order, frequency as the number of orders, and monetary value as gross profit across all orders.
4. Compare the averages, and the share of each group with a second order bought without a code.
5. Put both groups through the LTV calculator to see what each is worth after acquisition cost.

Some customers who used a code would have bought anyway, and some full-price customers came from referrals that were always going to be better. So the comparison isn't a controlled experiment. It shows which customers your offer attracts, which is enough to decide whether to change the offer.

One brand I worked with, a skincare brand doing about $18M a year, had run 15% off the first order in its welcome popup for three years. We replaced it with $20 toward the second order, good for 60 days. First-order conversion from the popup dipped very slightly. The share of new customers who came back within 90 days went from 9% to 32%, and the brand recovered the 15% it had been giving away on every first order. Email signups were the founder's biggest worry, and they barely moved. Leaving subscribers out, that cohort has earned three times what it cost to acquire.

Why does value per customer matter more as ad costs rise?

Because it's the lever you control. You can't set the price of a Meta impression. You can decide what a customer is worth once you have them. A brand that earns more per customer can keep buying traffic when costs climb, almost regardless of what acquisition costs, while a brand living on first orders gets squeezed out.

High value per customer also makes growth a choice. If you decide not to push for a season, you can pull back on spend and coast on profitable repeat orders from customers you already own. A brand built on discounts can't, because its customers only come back for the next code.

Start with the offer. Pick one of the five above, put it where your discount code is today, and measure the second-order rate for 90 days. To choose, look at the customers who already come back. They'll tell you what's worth rewarding.

Questions brands ask
about offers that
aren't discounts.