Did that promotion work? How to measure a coupon's signups and first payments in Recurly
A discount can fill your signup chart and still lose money. Here are the questions that tell you whether a promotion brought customers worth having, within days of launching it.
You launch a promotion: 50% off the first three months, a free first month, a code for a podcast audience. Signups jump. Everyone's happy. Three months later, half of them are gone and the other half are paying full price for something they never really wanted.
The time to find out whether a promotion is working is the first week, not the third month. Here is what to look at.
1. Who used the code?
"How many new subscriptions used coupon SPRING50 in the last 7 days, by plan?"
Compare that with how many new subscriptions came in without it over the same days. A promotion that mostly discounts customers who would have paid full price anyway isn't bringing growth; it's giving money away.
2. Did their first payments go through?
"Of subscriptions that used SPRING50 this week, how many had a successful first charge?"
A discount aimed at the wrong audience shows up quickly as declined first payments, particularly on free-trial or very cheap first-month offers, which also attract people testing stolen cards.
3. Where did they come from?
"For subscriptions using SPRING50, which traffic sources sent them?"
If the code was meant for one partner's audience and it's arriving from coupon-sharing sites, you know before the budget runs out.
4. Are they asking for refunds already?
"How many refunds went to accounts that used SPRING50, and for how much?"
Early refunds are an early warning that the offer promised something the product doesn't deliver.
5. Keep watching, automatically
"Alert me if more than 100 subscriptions use SPRING50 in an hour."
A runaway code, posted somewhere you didn't intend, can burn through a promotion's budget in an afternoon. A watch tells you while you can still do something about it.
The long view
Whether promoted customers stay is a question for renewal day. Ask the same questions again as their first full-price renewal arrives: how many renewed, how many declined, how many cancelled. For month-by-month cohorts, keep your history in a warehouse, or let Warehouse Sync load it into your own BigQuery.
Retrixus answers all of the above from live Recurly data, in Claude or ChatGPT. Start free on one Recurly site.
Try it on your own Recurly data
Start free on one Recurly site. Connect Claude or ChatGPT and ask your first question in about 15 minutes.
Start free