Selling the subscription, not the discount

Discount ads brought one-time buyers who cancelled after the first bag. Selling the routine instead kept seven in ten subscribers past three months.

Subscribers kept after 3 months

52% → 71%

Over 6 months

Past results, not a promise.

Client

Northroast

Coffee subscription

Channels

Meta

Monthly spend

$36k

The brief

Northroast roasts in small batches and ships a bag every two or four weeks. Their best-performing ads, by the numbers in Ads Manager, were discounts: 30% off the first bag. They brought orders cheaply. But half of those subscribers cancelled before the third delivery, and the ones who stayed came from somewhere else entirely.

Dana, the co-founder, asked a simple question: which ads bring the people who stay?

What we made

We matched three months of orders to the ads that brought them and found the stayers had clicked on anything that showed the routine: the bag on the counter, the morning cup, the price per cup. So we stopped selling a discount and started selling a habit. "Your 6am, sorted for $1.10 a cup." became the first of twelve ads built on cost per cup, skip-anytime and the roast date on the bag.

We kept a small discount running for comparison. Customers from the routine ads cost 18% more to win, but 71% of them were still subscribed after three months, against 52% before. Over a year that's worth far more than the cheaper first order.

What didn’t work

"Coffee, now 30% off." still had the cheapest first purchase in the account, and we switched it off on day three anyway, because almost none of those buyers made it to the second bag. We also tried a roaster-at-work video ("Roasted Monday, yours Thursday"). People loved it in the comments, but it never sold, so it now lives on Northroast's organic Instagram.