Retention Metrics: Start With Days Between Orders
The retention metrics an early-stage D2C brand or app should track first: days between orders, buying across categories, returning visitors, trial to habit.

Updated October 2026.
An early-stage D2C brand needs three retention metrics before any other. The first is the median number of days between a customer's first and second order. The second is the share of repeat customers who buy from a category they haven't bought before. The third is how many returning visitors go on to buy. Lifetime value can wait until you have a year of orders. For an app, the early numbers are different: onboarding completed, first task done, first task repeated, trial to paid, and how many users form a habit.
I'm Raviraj Ramavarapu, founder of Compound Lab, a retention agency for D2C brands and consumer apps. When founders ask me how they'll know retention work is succeeding, I give them these numbers before I suggest a single campaign. The vendor lists of 14, 17 or 19 retention metrics on page one of Google are written for companies with years of data. A brand in its first two years needs fewer numbers, read more often.
Which retention metrics should an early-stage brand track first?
The gap between first and second order, buying across categories, and returning visitors who convert. Each one answers a different question, and each points to a different fix.
| Metric | The question it answers | How to read it |
|---|---|---|
| Median days from first to second order | How fast do buyers come back? | Lower is better. Watch it by monthly cohort |
| Share of repeat buyers who add a new category | Do customers trust you beyond the first product? | Higher is better, and it lifts order value too |
| Returning visitors who buy | Are people who came back turning into customers? | Read it next to cart and browse abandonment |
| Lifetime value | What is a customer worth over time? | Wait until you have at least a year of orders |
Shopify covers some of this out of the box. Its customer cohort analysis report groups customers by the date of their first order, and its returning customer rate counts customers with an earlier purchase as a share of everyone who ordered. What it doesn't give you is the gap between consecutive orders. You build that yourself from an order export, which takes an afternoon with the sheet in my cohort analysis guide.
Why is the gap between orders the first number to track?
Because every later order hangs off it. If a customer's second order comes sooner, their third and fourth move earlier too, and more of them fit into the same year.
On a recent call with the founders of a fashion brand, it was the first thing I said I'd look at. Not the repeat rate, not the CRM, the number of days between the first and the second purchase. Say the median is 33 days. The job is to bring it down to 27 or 28.
That sounds small. Run the arithmetic for one customer who keeps the same rhythm all year.
| Median gap | Orders in a year, counting the first | Change |
|---|---|---|
| 33 days | 12 | Starting point |
| 27 days | 14 | Two more orders a year from the same customer |
That's the domino effect. A shorter first gap usually shortens the ones after it, because a customer who comes back quickly once has started to make you a habit. The money also arrives sooner, which matters more to an early-stage brand than to anyone.
Use the median, not the average. A handful of customers who return after eight months will drag an average up and hide what most people do.

How do you shorten the gap between first and second order?
Reach the customer just before they would have come back on their own, with a reason that fits what they bought. Most brands send their second-order nudge too late, or send the same discount to everyone.
- Time the nudge to your median, minus a few days. With a 33-day median, the first second-order message goes out around day 25 to 28, not day 40. What to send in the weeks before that is in my post-purchase flows.
- Get the first order right. Tracking updates, care and wash instructions, a note on fit. A customer who had a good first order comes back sooner than one who needed support.
- Offer the next item, not the same item. For a refill product the same item is right. For fashion or home, the next order is a different piece that goes with the first.
- Use the channel they actually read. In India that's often WhatsApp for the nudge and email for the detail, with consent for both. Check your WhatsApp delivery rate before blaming the message.
- Hold back a control group. Keep a small random slice out of the journey, so when the median drops you know the journey did it. I explain how in global control groups.
How do you measure product affinity across categories?
Count how many repeat customers buy from a category, material or range they haven't bought before. It tells you whether customers trust the brand or just one product.
Take a fashion brand that sells shirts. A customer buys a linen shirt. If the second order is another linen shirt, that's a repeat purchase, but the brand has only sold one idea twice. If the second order is a muslin shirt and the third is jacquard, that customer now trusts the brand's taste, not just one fabric. That trust is what grows order value and makes the next launch easier to sell.
| Measure | How to calculate it |
|---|---|
| Cross-category repeat rate | Repeat customers who bought a second category, divided by all repeat customers |
| Categories per customer | Distinct categories bought, averaged across repeat customers |
| First category to second category | Which first purchase leads to which next one, from your order data |
The last one is where the action is. Group buyers by what they bought first and look at what they bought next. The next product each group points to becomes the offer in their second-order journey. I go deeper on mapping it in replenishment and cross-sell.
Should returning visitors count as retention?
Yes, for an early-stage brand. Someone who came to your site, left, and came back is showing the same interest a repeat buyer does, just earlier. If they keep coming back without buying, the problem is on your site or in your follow-up, not in your ads.
Google Analytics 4 counts returning users as people who started at least one earlier session. Compare how many of them buy against how many new visitors buy, week by week. If returning visitors don't convert much better than new ones, look at what happens after a product view or a cart. My guides to browse abandonment and WhatsApp cart recovery cover the follow-ups that bring them back.
When should an early-stage brand start tracking lifetime value?
Once you have at least a year of orders and enough repeat customers that the number doesn't swing with one good month. Before that, a lifetime value figure is mostly a guess built on a few months of data.
Klaviyo's own rules make the point. Its predictive analytics, which estimates each customer's expected next order date and lifetime value, needs at least 500 customers who have ordered, 180 days of order history, recent orders, and some customers with three or more orders. Many brands in their first year don't meet that. When you do, the LTV to CAC ratio guide shows how to use the number.
Which retention metrics matter for an app?
For an app, retention is a chain of first steps, and each one has its own number. Are people finishing onboarding, doing the first task, coming back to do it again, using the trial fully, paying, and building a habit?
| Stage | The metric | What it tells you |
|---|---|---|
| Onboarding | Share who complete onboarding | Whether the flow is clear |
| First task | Share who do the core action on day one | Whether the app made the next step obvious |
| Repeat task | Share who do it again within 7 days | Whether the first task was worth repeating |
| Trial use | Days active during the trial | Whether the trial shows the value before it ends |
| Trial to paid | Share of trials that convert | Whether the paywall and the trial worked |
| First habit | Share who reach your habit mark, such as 3 active days in week one | Whether the app is becoming routine |
| Streak | Users on a streak of 7 days or more | Whether the habit is holding |
Streaks are worth tracking even if your app doesn't show them to users. Duolingo found that learners who reach a 7-day streak are 3.6 times more likely to complete their course. Your habit mark will be different, so find it in your own data. Look for the number of active days in the first week after which most users stay.
For the first three stages, see getting new app users to their first action. Trial to paid, with the India benchmark, is in paywall conversion, and daily use is measured in DAU/MAU benchmarks.
How often should you review retention metrics?
Weekly for the numbers you're actively trying to move, monthly for the rest, and always by cohort rather than as one blended figure.
A blended number mixes customers from last year with this month's first-timers. A new cohort that's doing worse can hide behind loyal old ones. So read the gap between orders, the cross-category rate and the app stages for each month's new customers, and ask one question. Is this cohort better than the one before?
Where I fit
Compound Lab works with D2C brands and consumer apps on everything after the first order or install. That includes setting up these numbers, finding the one that's furthest off, and building the journeys that move it. If you're early and not sure which retention number to watch, book a 30-minute call and we'll pull your gap between orders from your own data.