Naked Wine: Customer Lifecycle

Can we walk through that? Let’s say you walk into Naked, first few weeks, what was the plan? How did you segment the customers, to begin with, for example?

Just to take a little step back, just before we dive into the detail on that, Naked, as a subscription business, is very much based around customer economics. The unit of analysis, at Naked, is customers. To contrast that, the unit of analysis at Majestic was very much the store. Was the store performing? Was it in the right location? Did it need more footfall? Ultimately, that became a sales and a like for like measure. At Naked, as a subscription business, and one of the early online subscription businesses, the economics were very much centered around the customer. You saw that in how they were set up and how they thought about the customer lifecycles and how you balance the growth with the customers you lose, along the way. It was very much set up around this thinking about, I have a customer, but what I really want is a mature customer, so how do I get to the point of a mature customer?

When you do these kind of lifecycle economics, what you find is that you can never be super precise around it. Firstly, you are talking about lifetime values and who knows what is going to happen in the future. I would imagine that, with coronavirus, all of Naked’s lifetime values took a massive jump upwards because customers who might have got a few of their bottles of wine from the supermarket, every now and again, and used Naked for special occasions, are suddenly much more engaged and very happy customers.

When you do these lifecycle economics, you are talking about a lifetime where you are probably going to retain a customer for 15 or 20 years, but you’ve only been operating for three or five years. It’s very hard to project what it’s going to be. You really have to work in broad brushes of, for example, just trying to move customers from one segment into another segment. For Naked, if you buy a 12-bottle case of wine, for a lot of people, that would probably last two to three months. For them, it was really about, what about the second case or are people contributing into their angel account for more than four months? That was the real defining point at which you say, the economics, in terms of the retention and things like that, get a lot better, so that’s how we should think about it.

The company was really set up about those mature angels. How do I get them to that point and how to I optimize for that? Interestingly, going back to the price discussion, if you are very competitive on price, especially on the intro case, you get a lot of customers on the top, but they just churn out faster because you are attracting customers who are there for the cheap wine and they’re not really buying into the story. Whereas, if you can work out a way to engage customers such that it identifies customers who really are interested in the wine, then those are the ones that you want to go after.

For a lot of these subscription businesses, it’s like a big sieve. You have a lot of customers come in the top and you give it bit of a shake and you’re really just looking for the ones that stay and become good customers. There is a real balance there of, what is the right metric to look at, to know if my customers are going to become these good customers? Obviously, voucher acquisition is very important for Naked. I’m sure everyone has seen their vouchers, which are very ubiquitous. But one of the questions is, if I sign a new voucher partner, if I start putting vouchers in a new magazine subscription or a new e-commerce retailer, what kind of customers will they drive? Are they going to drive good quality customers or poor quality customers? The differentials in those are pretty high.

There is lots you can do, in terms of trying to predict how good those customers are and there is stuff that you can know on day one. One classic example is the credit card that someone uses to pay for their wine is a really big predictor of how long they are going to go on. Amex, obviously, being a predictor of higher value; someone using a pre-paid Mastercard, very likely to be very low value.