We cut our discount rate from 33% to 3%
Happy Thursday!
Earlier this week I made a LinkedIn post about how we reduced discount rate in Q1 this year.
We thought we'd been doing too many promos for a couple of years. Over time it became a crutch, and price is the easiest lever to pull.
The bet we made this year was simple: cut the everyday discounting, use the time the team got back to focus on fundamentals, and give it six months to see if it drove improved profitability. We still do the big holidays. Father's Day, Black Friday, etc. What we decided to cut was the everyday cadence.
We went all in, and our average discount rate of 33.7% in Q1'25 went down to 2.9% in Q1'26.
The bet was on quality of revenue
Revenue that needs a discount isn't as valuable as revenue that doesn't. (would love to hear if anyone disagrees with this... email me back if so)
A few weeks back I wrote Discounts aren't a growth strategy, and that was the philosophy. Q1 was our first real test of that theory.
We knew unit volume would come down when the discounts stopped. We didn't know by how much, though. Our bet was that the margin we kept on every full-price order would outweigh the units we lost, and it did.
Amazon's surprise
We thought cutting promos on Amazon would meaningfully hurt topline. The opposite was true. Revenue went up YoY and average sales price per unit increased 31%. People bought fewer items but spent more on each one. When we stopped training them to wait for a discount, we were surprised to see dollar demand didn't dip. This single change meaningfully improved contribution margin from Amazon as a channel.
What we did instead of promos
When you're running a promo every week, a lot of the team's time goes into planning it. Briefing creative, building landing pages, scheduling emails, and reporting on results.
Take that out and the team has time for the fundamentals. Better product pages, sharper messaging, and better evergreen testing cadences. The product page rebuild I wrote about in How we cut our Meta CPA in half without touching Meta came out of this same window.
What it did to profit
Profit for Q1 came in significantly better than prior year. A few things stacked to get there.
It started with the realized sales price. When you stop giving up a third of every order as a discount, you limit margin compression and give yourself some breathing room. Gross margin improved year over year, contribution margin grew, and the fundamentals the team focused on compounded the results from there.
A better-converting site made each ad dollar more efficient, so ad spend as a percentage of revenue decreased 5+ points. Some of the volume we used to buy with promos ended up being retained through pages that sold better during evergreen periods. Underneath all of it, opex came down meaningfully year over year.
This is what the quality-of-revenue trade looks like in practice. It led to better margin, better customers, better staying power, and stronger brand. We took a calculated risk, and so far, it's paid off.
What to do next
Run the breakeven math on your own discounting. You need three numbers: average order value before discounts, your realized discount rate (gross sales minus net sales, divided by gross sales), and your variable cost per order (COGS, shipping, fulfillment, ad spend, etc.).
Contribution per order today is AOV × (1 − discount rate) − variable cost. Contribution per order at full price is AOV − variable cost. Divide the first by the second. That's the share of volume you'd need to keep for cutting discounts to break even. Keep more than that and you're ahead.
For example: at a 30% discount rate with variable costs at 40% of AOV, you break even keeping just half your units. Most teams assume they need to keep far more volume than the math actually requires. That gap is the whole opportunity.
If you've cut discounting and seen a different outcome, reply and tell me why.
Have a great weekend, and see you next week!
Kyle
P.S. Did you find this useful? If so, pass it on or send them to kyleyeoman.com to subscribe.