Answer
Do abandoned cart emails actually add revenue?
They may, but attributed revenue can't tell you how much. Your flow report credits the email for orders that follow it, and some of those shoppers would have come back anyway. To find out what the flow adds on your store, run a holdout: keep a random share of shoppers out of the flow and compare total orders.
October 31, 2026
They may, but attributed revenue can't tell you how much. Your flow report credits the email for orders that follow it, and some of those shoppers would have come back anyway. Only a holdout on your own store can show what the flow adds.
Why isn't attributed revenue enough?
Klaviyo's 2026 benchmark report puts abandoned cart email at $6.77 of attributed revenue per recipient on average, and $13.70 for the top 10%. That makes it a strong performer on the metric, and the figure is useful for comparison.
It also has a built-in bias. People who just added something to a cart were already interested. Some would have bought with no email at all, and the report still credits the flow when they click and order.
| Attributed revenue | Incremental revenue | |
|---|---|---|
| The question | Which orders followed this message? | Which orders happened because of it? |
| How you get it | A report in your email tool | A comparison with a held-out group |
| What it's good for | Comparing flows and spotting problems | Deciding what the flow is really worth |
How do you test it?
- Split people entering your abandoned cart flow at random, and keep one group out of the emails.
- Count orders per person in each group, from every source and channel.
- Set the length and group size in advance, and don't stop at the first good-looking result.
- Check other flows, because a held-out shopper may still receive a checkout email or a campaign.
The result might be smaller than your report suggests. It might also show that the flow pays for itself comfortably. Either answer is better than an assumption, and it can differ from store to store. Send timing and discounts are worth testing the same way, since a discount costs you on orders you'd have had anyway.
What are the two separate questions?
"Does this flow add revenue?" is one question. "Does this button destination do better than my current one?" is another, and it needs its own test.
reOtter's cart recovery page has a built-in holdout for the second. It sends a share of shoppers to your current destination and compares orders. It doesn't measure whether the flow itself adds revenue, so you'd still need a flow-level holdout for that. No brand has run reOtter's test to a result yet, and reOtter makes no lift claim. Measuring cart recovery with a holdout explains the setup.
Key takeaways
- Attributed revenue credits orders that follow a message, and some of them would have happened anyway.
- A random holdout with all orders counted is the way to estimate what the flow adds.
- Testing a flow and testing a landing destination answer two separate questions.
See how reOtter restores the cart → Abandoned cart recovery
Frequently asked questions
- What's the difference between attributed and incremental revenue?
- Attributed revenue is what a report credits to a message based on rules such as a click before the order. Incremental revenue is what wouldn't have happened without the message. Only a comparison against a group that didn't get the message can estimate the second.
- How do I run an abandoned cart holdout?
- Split shoppers entering the flow at random, keeping one group out of the emails. Compare orders per person across both groups, counting orders from every source. Choose your length and group size before you start, and don't stop when the numbers first look good.
- Can reOtter's holdout tell me if my cart flow adds revenue?
- It can't answer that question. reOtter's holdout sends a share of shoppers to your current destination and compares orders, which shows whether the page beats that destination. No brand has run it to a result yet.