The welcome flow discount is costing you more margin than it earns

Nobody chose that discount, it shipped with the template

The first email in that client's flow led with a discount code straight out of the box. We asked where the number came from. Nobody could tell us. It had been there since launch, copied from a Klaviyo template two agencies ago.

Every person who touched the account assumed the code was load bearing. It wasn't. That's the ordinary way a welcome flow rots: not through a bad decision, but through a decision nobody remembers making.

This isn't a discounts are bad argument. The habit worth breaking is opening the flow with a blanket code because that's what welcome flows do. It's a toll, and nobody's checked whether you still need to pay it.

Work out which job the discount is actually doing

A welcome discount can only do one of two jobs. It either wins a customer who was never going to buy without it, or it hurries up somebody who was already going to buy. Different jobs, different economics, and most brands never separate them.

The customer who needed the code is price sensitive. They'll be price sensitive again at the second order and every sale after it. Repeat purchase gets decided in the gap between order one and order two, and a discount led signup starts that gap on the back foot.

The one who was already sold is quietly costing you money. If your product has real demand, a chunk of your signups handed over their email because they want the thing. The code is margin given to people who never asked for it, on every one of those orders.

Compare welcome flow order value against the rest of the site

One check tells you which job your code is doing, and it takes an afternoon. Pull the last ninety days of welcome flow orders and set their average order value against the sitewide average. The gap between those two numbers is your diagnosis.

If welcome flow order value sits well below the site, the offer's pulling in bargain hunters instead of converting fence sitters. If the two are roughly in line, the discount probably isn't why most of those people bought.

We read a report like that in revenue per recipient, not the conversion column, and that's how we judge every flow across the accounts we run. A flow that converts brilliantly while dragging order value down isn't a top performer. It's a leak with good optics.

Run a value led first email against the code before you decide

The test we ran wasn't remove the discount and hope. It was a straight split, same structure, same timing: one arm led by the code, the other led by value. What the product does, the proof it works, the founder's reason for making it.

The value led arm converted at a lower rate. That's expected, because some people genuinely won't buy without an incentive. But everyone it did convert paid full price, and the recovered margin outran the lost conversions. Fewer people through the door, more money staying per person.

A conversion rate on its own tells you nothing about profitability, and on a dashboard the discount arm looks like the winner. Run both arms through margin per send and the value led one can still come out ahead, because it isn't subsidising buyers who'd have paid full price anyway. Give the test a full flow cycle, sixty to ninety days. Then judge it on margin banked.

Know when the discount genuinely earns its place

Look, this only works if your product and your proof are ready to do the job the code used to do. Pull it from a brand people don't trust yet and you'll wreck conversion, and we'd tell a client that to their face before running the test.

• You're a genuinely new brand with no repeat history and no reviews at volume, so the code is doing trust building work a value led email can't do alone yet
• Your order value is so low the code barely dents margin per unit, because a small amount off an impulse buy behaves differently to the same cut on a considered purchase
• Your category expects a first purchase code and every competitor runs one, so going without leaves yours the only checkout with an empty box people went looking to fill
• You've got inventory or cash flow reasons to prioritise volume over margin for a set window, which is a real decision as long as somebody actually made it

If none of those apply, and especially if you've already got returning customers, decent reviews and organic demand, you're very likely the brand at the top of this page. The code runs because it's always been there. Not because anyone tested whether it needs to be.

Rebuild the flow around the reason people signed up

Pull the code and the rest of the flow has to earn the sale, because you've removed the thing doing the persuading. This is where the welcome flow stops being an introduction and starts being a selling document, and where it overlaps with the conversion testing we run.

Email one answers the question the signup implied. Somebody who joined from a blog post about ingredients wants ingredients. Somebody off a TikTok link wants the thing they saw, fast, with proof it's real. Generic welcome copy only ever worked because the code underneath was doing the lifting.

Email two carries your best proof. Reviews with specifics instead of a star graphic, or a founder explaining the problem the product solves. Email three is the last full price push before you decide whether a smaller, later incentive is worth reintroducing. If it comes back, it comes back as urgency for fence sitters, never as the headline of the flow.