When the Customer Changes and the Marketing Doesn't

Late last year I started an engagement with a consumer goods brand doing about twenty-one million dollars a year through Shopify. The founders had built the business over several years of focused digital marketing, running ads and content around a specific customer profile that had carried them to that point.

When we pulled eighteen months of their order data together for the first time, we could see that the customer base had shifted. The profile the marketing was built around wasn’t the profile that was actually buying. The vast majority of buyers had purchased exactly once, which is not unusual for consumer goods at that stage, but steeper than their unit economics could absorb. The data had been sitting in the Shopify account the whole time. Nobody had pulled it all together to look at the full picture.

I’ve seen some version of this at most mid-market companies I’ve worked with. A marketing stack, customers served for years, and a gap between who the brand thinks is buying and who actually is.

Where growth outpaces the data practice

Companies making it to this size did it by investing in the things that obviously create revenue: product, marketing, acquisition. Data infrastructure is an investment in decisions you haven’t made yet, and the cost of making those decisions on instinct only shows up later.

There isn’t a clean revenue threshold where this always happens. It’s more about when the founding team can no longer hold the full picture through conversations and instinct alone. At some point there are too many customers and campaigns for anyone to carry it, and something about the business shifts without anyone noticing until the numbers stop working the way they used to.

Starting from the decision, not the data

The most useful first move on a new engagement is to understand what the leadership team is trying to decide. Not what data they have, but what call they’re trying to make this quarter that they don’t have good information for. At this brand, the question was whether their current marketing approach was reaching the customers who were actually converting.

That framing shaped everything. The first month was in service of answering that one question.

What the first ninety days looked like

Week one was pulling the full order history and running an RFM cohort view the team had never seen. Week two, we partnered with a third-party consumer data provider and overlaid about fifty percent of the Shopify customer base with demographic and behavioral attributes. That overlay let us see customer profiles quickly and clearly, and it showed the founder that the average buyer was more than a decade older than the customer his content strategy was built around.

That was the moment the engagement shifted. The marketing had been optimized for one demographic. The revenue was coming from a different one. Nobody was wrong. The business had grown, and the customer had changed with it, but the marketing hadn’t caught up.

Week three was back-of-envelope math on a retention play. Five percent of roughly four hundred thirty thousand one-time buyers at their AOV pencils out to about a million dollars a year, more if repeat cohorts have better LTV than first-purchase cohorts, which they usually do. Weeks four through six were building out the segmentation to help the team see which customer profiles, channels, and geographies were worth doubling down on.

A month of focused work to answer one question. That’s the kind of foundation that opens up the next set of conversations.

Where the work gets harder

I don’t think I’ll ever fully calibrate for this part. Even when the data makes the answer clear, the team at the top of the company often needs time to act on it. At this brand, the data said the marketing was aimed at a customer the brand was not selling to. It said the retention play was worth real money. The leadership team heard all of that and, for a stretch, kept running the playbook that had gotten them the last two years of growth.

I don’t think that is stupidity. The instinct that got them to twenty million is the same instinct the data is pushing against, and asking a founder to override that on a cohort analysis is asking them to distrust the thing that worked. The diagnosis earns you the right to have the harder conversation, and the harder conversation is where these engagements live or die. I have been on both sides of that, including cases where I did not earn enough trust fast enough and the window closed.

What the team has now is a defensible story about which customers to pursue, which channels to invest in, and where the retention opportunity lives. That’s the one thing they didn’t have going in.

What the first twelve weeks are for

The sequence matters. Week one to week twelve in the right order, starting from the question the leadership team is already wrestling with. The people, the conversations, and the willingness to pick one question you can answer in two weeks instead of ten you’d like to. I think that’s why “what are you trying to decide” is the right opening question. It gets you to the work that matters fastest.