How Do I Test Upsell Offers in Subscription Plans?

The subscription sign-up is a uniquely high-intent moment. A customer has chosen to buy from you on a consistent basis, but hasn't placed that first order yet. That's what makes it the best opportunity to grow the order, and the easiest one to mishandle. In this post we'll walk through two types of offers worth testing at that moment, why you should test whether any offer helps before testing which one, and what to actually measure when you run it.

Two Ways to Grow the Order at Subscription Sign-Up

There are two clear types of offers worth testing at the moment someone subscribes. (Technically, both are cross-sells, you're adding a different product, not upgrading the one they picked. But "upsell" is how most brands and apps refer to this, so we'll use the terms interchangeably here.)

The first is a subscription add-on: a second product they'd receive and be charged for on an ongoing basis alongside their main subscription. A supplement brand might offer to add creatine or fish oil to a monthly protein order. A pet food brand might suggest dental chews alongside the regular kibble. These aren't one-time purchases... they fold into the recurring relationship.

The second is a one-time complementary item: something that doesn't recur, but makes the subscription more useful or complete. A razor subscriber might add a travel case. A coffee subscriber might add a seasonal roast. A skincare subscriber might pick up a facial tool they'd only need to buy once.

Both lift the value of that first order. The difference is that add-ons keep compounding. A second recurring product means more revenue on every renewal, not just the first one.

Why You Test This Instead of Just Adding It

An offer that fits can feel like a helpful recommendation. An offer that adds friction at the wrong moment doesn't just lose the add-on sale. It can cause someone to abandon the subscription entirely.

That's the risk experienced operators already know about. The question isn't whether the upsell opportunity is there. It's whether your upsell is helping or quietly cutting into subscription completion rates. That's not something you can assume. It's something you test.

Start With One Question

Before you decide what to upsell, ask yourself one thing: does showing any offer here help or hurt?

That's the first test worth running. Not "which product should I show?" Not "should I discount it?" Just: upsell present vs. no upsell. Does subscription completion rate hold when you add an offer? Does it drop?

If it holds, you have runway to optimize. If it drops, you've learned something important before building anything else.

IntelliJAMS EP 018: Juan Cruz Giusto and Alex on what's actually worth testing in a subscription model.

Then Test the Offer Itself

Once you know an upsell can live in the flow without hurting completion, you can start testing what the offer should be.

For subscription add-ons, look at your order data first. Are subscribers placing separate one-time orders for complementary products outside their subscription? That's your first candidate. A customer who subscribes to protein and separately reorders a pre-workout every few months is telling you something.

For one-time items, look for the natural complement to what they're subscribing to. The product that makes the subscription more useful from day one. The shaker bottle. The pill organizer. The tool that completes the routine they just committed to.

Start with one product at a time. It gives you cleaner signal on what's actually working. That said, your traffic volume should frame how aggressive you can be. Higher-volume stores have more room to run broader tests without losing statistical confidence.

What to Measure

Take rate on the upsell is the obvious thing to watch. But it only tells you whether people are clicking yes.

The metrics that actually matter are subscription-specific.

Watch subscription orders per visitor and subscription % of orders alongside take rate. If either drops in the upsell variant, the offer is creating friction you won't see in the upsell numbers alone. You might be winning the add-on and losing the subscription.

Subscription profit per visitor tells you whether the offer is worth running at all. Revenue going up doesn't mean you're better off if discounts and fulfillment costs eat the margin.

These and other key subscription performance metrics are available on Intelligems out of the box.

Then there's the longer view. Intelligems' post-test metrics feature lets you measure what happens after the test window closes: reorder rates, retention patterns, customer quality. For subscription brands, the real question isn't just what happened during the test. It's whether the customers acquired through the upsell variant kept subscribing.

The Compounding Part

Once you find an offer that works without hurting completion, every renewal that follows is more valuable than the one before. That's the unlock. The brands that pull ahead don't stop there. They keep testing: offer type, product selection, discount depth, framing. Each iteration compounds on the last.