Everything we sell online at Buon Cibo, from preorder dinners to event tickets to prepared food, is tied to a customer’s email address. So I pulled six months of online orders, took out cancelled, refunded, and zero-dollar orders, and looked at what people actually did, customer by customer.
Most owners never look at their business this way. They look at sales per day. Looking at it customer by customer changed how we think about almost everything, and it’s the first analysis we’d run for any owner we work with. Here’s what we found.
The second order is the one that matters
Our weekly dinner program is the thing most people buy first, so it’s where we measure repeat business most closely. Of the customers who had a full 90 days to come back, 57% placed another order within those 90 days.
| Window | Reordered |
|---|---|
| Within 30 days | 34% |
| Within 60 days | 45% |
| Within 90 days | 57% |
Across everyone who has ever bought that dinner, the number is 39%. That’s lower because it includes people who placed their first order last week and haven’t had time to come back yet.
Both numbers are true. They just answer different questions. This is where most repeat-rate claims go wrong, so if anyone ever tells you their repeat rate, ask two things: out of whom, and within how long?
Repeat orders are most of the money
More than half of what the weekly dinner program brings in (53%) comes from orders placed after a customer’s first order of it. The first order gets someone in the door. The orders after it are the business.
Across all our online orders, customers who bought on more than one day account for 58% of sales. That’s in the same range Olo found across more than 100 million restaurant guest records, where repeat guests drive about 60% of revenue. The first purchase is the beginning, not the result.
Our best customers kept going
More than one in ten of our online customers have now ordered on four or more different days. That’s the group we watch most closely. Paytronix, which runs loyalty programs for more than 800 restaurant brands, found that guests who reach a fourth visit come back 95% of the time, compared with less than half after a first visit. We don’t claim that rate for our own customers, but it’s why getting someone to a fourth order is the goal. I wrote more about that in customer lifetime value and repeat rate for restaurants.
New customers kept coming, too
Repeat business only works if new people keep walking in the front door. Comparing the most recent 28 days with the 28 before, the number of people placing their first online order with us rose 127%, and paid orders rose 87%.
One honest caveat: “first online order” isn’t always “first visit.” Some of those people had probably bought something at the counter before. What it does show is more people moving into the part of the business we can reach again.
The weekly dinner was the front door. For nearly nine in ten of its customers, it was the very first thing they ever ordered from us online.
People who buy one thing buy another
About one in five of our program customers have bought from more than one program, and they account for about a third of program sales. Half the customers of our other meal program have also bought the weekly dinner, and for most of them the dinner came first. Every program ends up being a way to introduce the next one.
- Export every order with the customer’s email or phone number, the date, and the amount.
- Take out cancelled, fully refunded, and zero-dollar orders.
- Count customers, not orders, and only count a repeat when it’s on a later day.
- Measure repeat rate at 30, 60, and 90 days, counting only customers who’ve had the full window.
- Work out what share of sales came from orders after each customer’s first.
- Check which products or programs the same customers buy together.
- Keep a list of customers with four or more orders, and watch it grow.
If your system can’t export customers at all, that’s the first thing to fix. You can’t grow repeat business you can’t see.
What we took from it
- The first order is an introduction. The second one is what makes a regular.
- Always say who you’re counting and over what window, or the number means nothing.
- Most of the revenue from a good program comes after the first purchase.
- One simple, well-priced program can be the front door for everything else.
- Customers who buy across offerings are worth far more than their numbers suggest.
How we counted: Buon Cibo’s online orders over about six months, with customers identified by email address. Cancelled, fully refunded, and zero-value orders are excluded. “First-time online buyers” means first seen in the online orders, not necessarily first-time guests. Sales are net of discounts and refunds, before costs. These are one business’s results, not a promise of what any other business will see.