Case studies: Multi-state cannabis retailer

Acquisition got the credit. Retention did the work.

Cannabis retailer · Five states · Display advertising, January to August

Let's Talk

5.09x

Return on ad spend from new and won-back customers

35.6%

Were first-time customers

26.5%

Were lapsed customers who came back

The campaign

Display advertising across a multi-state store footprint

This retailer operates stores across five states with sales data flowing back to the campaign. Chains at that size face a specific problem: it is easy to prove advertising reached people, and hard to prove it changed what they did.

MediaJel ran display advertising for roughly eight months, then matched it against the retailer's own order records rather than counting clicks. Every customer was sorted by what they had bought before the campaign began, using two years of purchase history.

A comparison group was also built from existing customers the ads never reached, picked so they had last shopped around the same time as the customers who did see the ads. That made it possible to compare how many of each group kept buying.

Who the customers were

Three different kinds of customer, in similar share

Separating the won-back group from the regulars meant checking every customer against two years of prior orders. A shopper returning after months away places an order that looks no different from a loyal customer's, so without that step this group would have been counted as ordinary repeat business.

RETENTION

Customers who saw the ads were far more likely to still be buying

This is the strongest finding in the campaign, because it is the one with a comparison group behind it. Two thirds of the customers who saw the ads were still buying at the end. Fewer than half of the comparable customers who saw none of them were.

19 points

Of first-time buyers the campaign reached bought the brand again within 60 days

36,000

Customers included in the comparison, across both groups

How fast they acted

A quarter of orders landed the same day or the next

Two thirds of attributed orders arrived within a week. The remaining third came in over the following three weeks, which is the part a short attribution window would miss entirely.

Find out what your advertising is actually changing

Reach is easy to prove. Whether it changed what customers did takes a comparison group and your own order data. Let's talk about building that.

Start the Conversation