Measuring DTC Customer Lifetime Value: Metrics & Strategies

I stopped tracking lifetime value and started tracking 180-day contribution margin per acquired customer (CM180).

Men’s rings are a considered, low-repeat purchase. A three-year LTV model gives me a number I won’t collect until 2029, useless for deciding whether to spend $40 on a click today. Six months captures nearly all the revenue a ring customer will ever generate, so CM180 is close to true lifetime value but I can act on it in one quarter instead of three years.

The math: cohort revenue through day 180, minus COGS, shipping, payment fees, and returns. Then divided by customers acquired in that cohort.

The rule I run against it: if CM180 / CAC is under 1.5, the channel dies. That one ratio is what told me cold paid social doesn’t work at our ~$150 AOV. It never cleared 1.0 and moved budget into organic search and post-purchase retargeting, where it does.

Most DTC operators measure LTV over a window longer than their ability to react to it. Shorten the window until the number becomes a decision.

The key problem is that most DTCs focus on blended LTV, which obscures the erosion of margins from continuous retargeting. The one metric you should be looking at is Zero-CAC Repurchase Rate (ZRR). This is the % of first-time buyers who eventually make their 2nd and 3rd purchases via direct traffic, organic, or AI brand citations — aka, they never click a retargeting ad.

I’ve seen this metric applied in the wild by a mid-market DTC brand in the health/wellness space, and they started by telling me their 12-month LTV was $110, based on their analysis. However, their retargeting — which was happening at the bottom of funnel, at high frequency — was killing their margins. When we sliced the LTV by the acquisition/first-touch channel, the difference became clear. When acquired via interruptive paid social ads, the ZRR (90-day, for example) was 12%. When acquired via high-intent organic search and AI engine recommendations, the ZRR was 38% — reflecting the inherent trust and self-motivation their buyers had.

What they did with this data was shift their acquisition strategy towards getting dominant organic search visibility, as well as getting cited well in AI. The net effect was that the average LTV over 12 months directly attributable to first-touch acquisition went from $110 to $145 — with a contribution margin on the returning customers shooting up from like 15% to 41%.

To implement, simply set up your preferred analytics tool to cohort by first-touch acquisition channel, and then assess 90/180-day repeat purchase rates — then flag any channel that involves a retargeting click by excluding any subsequent purchases that happen with paid UTMs.

ZRR shows you which channels drive sticky customer acquisition and which ones rent eyeballs.

Tracked average order value for about two years before realising it told us almost nothing useful about customer quality, because a customer who bought once at $180 and a customer who bought six times at $40 looked very different on that metric, and the second one was considerably more valuable to the business.

Switched to measuring predicted 12-month revenue per customer, calculated at the 90-day mark based on purchase frequency and category in the first three months.

The 90-day window was specific rather than arbitrary. Purchase behaviour in the first three months was the strongest predictor of 12-month value in our category, with a correlation that remained reasonably consistent across two years of cohort data.

According to Shopify’s research on DTC customer behaviour, customers who make a second purchase within 90 days have roughly 5 times the lifetime value of single-purchase customers, which matched what we were seeing.

We use the metric to decide which acquisition channels are worth investing in and which aren’t, measured by the predicted 12-month value of customers they generate rather than acquisition cost alone.

Most people overcomplicate lifetime value. For a DTC store you rarely need a predictive model, a rough version tells you almost everything and you’ll actually use it.

The way I track it, average order value, times how often a customer buys in a year, times how many years they usually stick around. Simple, but it tells you what a customer is worth over time, not just what they spent on day one.

Where it matters is deciding what I can spend to acquire someone. If a first order is $50 but a customer’s worth $200 over two years, I can outspend competitors who only look at that first sale, they panic when acquisition costs more than the opening order, I don’t.

The metric I’d never drop is repeat purchase frequency, because it’s the part of LTV I can actually move. Get existing customers to buy one more time a year and you lift the whole base’s value without spending a cent on new traffic.

We calculate the DTC customer lifetime value through the analysis of the 12-Month Cohort Retention Curve, times average net subscription revenue per user – both are fully adjusted for gross margins and app store fees. The cohort-based methodology removes the normal seasonal increases in churn that are common with all subscription services; therefore, it allows us to accurately calculate our LTV on a margin basis as opposed to using a bloated historical average.

In addition to tracking the customer lifetime value, we also track and monitor the CAC payback period. The CAC payback period calculates how long after customers acquire their first subscription service until they have generated enough gross profit to cover the full cost of acquiring them. As such, we utilize the CAC payback period to continuously assess and adjust our advertising spend across various media outlets. In turn, we can reallocate advertising budgets only into those campaigns that achieve at least 100% capital payback within the timeframe of 60-90 days.

I look at lifetime value through trust, not just the first booking. In our storage and removals business, someone might use us once for a move, then come back later for renovation storage, or refer a family member who is downsizing. The metric I care about is repeat and referred enquiries after a completed job. It tells me whether the experience was strong enough to carry into the next stressful life moment. That is more useful than only looking at the value of one transaction.

The State of eCommerce in 2025!

The metric I care about with direct to consumer clients is the share of revenue coming from returning customers, tracked month by month against the same month last year.

Lifetime value itself is a modelled number and it flatters everybody. It gets built from an average order value and an assumed repeat rate, and if either assumption is optimistic the whole figure is fiction. Returning customer revenue share is a fact out of the accounts. If it falls while total revenue rises, the business is renting its growth from ad spend.

We use it to decide what acquisition is worth. A store where returning customers make up 29% of monthly revenue can afford to pay considerably more for a first order than one sitting near zero, because that first order is not the whole relationship. That single figure changes how aggressive we can be on paid search, and it is the number I ask for before quoting on an account.

It also catches problems a conversion rate hides. On one account the ratio slid for two quarters while everything else looked healthy. The cause was a fulfilment change that had slowed delivery times, and no marketing lever would have fixed it.

Pair it with gross margin per order after advertising. Between them, those two tell you whether the growth is a business or a habit.

Marketing Calendar: A Complete Guide

The one I actually watch is the 90-day repeat rate, not lifetime value in the classic sense. Real LTV takes months or years to know for sure, and by then it is too late to do anything about it. But whether someone comes back within about 90 days tells you almost everything early. It predicts the long-term value way before the long term actually shows up. So we use it as the steering wheel. If a channel or a promo brings in buyers who don’t come back inside that window, it doesn’t matter how cheap the first order was, those customers are usually worth less than they looked on day one. We shift spend toward the sources whose buyers repeat, and we build the early experience, the first follow-up, the second-order nudge, around pushing that 90-day number up. One more thing, look at it in cohorts, not as a blended average. An average hides the fact that a small group of repeat buyers is quietly carrying the whole business.

The State of eCommerce in 2026

Most DTC brands calculate LTV wrong. They look at historical purchase data and call it a day. When I was running my e-commerce brand, I tracked something more predictive: second purchase velocity.

Here’s what I mean. If a customer bought again within 45 days, their lifetime value was 4.2x higher than someone who waited 90 days for that second order. That single metric told us more about future revenue than any fancy cohort analysis. We obsessed over it.

The way we used it was ruthless. Every marketing dollar got evaluated against how many sub-45-day repeat buyers it generated, not just first-time customers. Our email sequences completely changed. Instead of generic “thanks for your order” messages, we built triggers based on product usage cycles. Skincare runs out in 6 weeks? Email goes out at day 38. Not day 40, not day 45. Day 38.

We also fired wholesale accounts that cannibalized our repeat rate. One big box retailer wanted to carry our hero product. Would’ve been a vanity win and decent revenue. But our data showed that customers who discovered us in-store had an 80-day second purchase window versus 42 days for our DTC channel. We walked away from the deal because the LTV math didn’t work long-term.

The mistake I see brands make through Fulfill.com is treating LTV as a reporting metric instead of an operating metric. It should change how you spend money today. When Nature Hills Nursery switched 3PLs through our platform, their faster shipping didn’t just improve satisfaction scores. It compressed their repeat purchase window by 11 days, which translated to 19% higher LTV across their spring cohort.

Track the time between first and second purchase. Everything else is just accounting.

The Complete 2025 E-commerce Manager Career Guide: From Entry-Level to Executive

I will give the uncomfortable version, because most brands I work with sell primarily on Amazon and simply do not own the customer. You get no email, no name, and no ability to retarget an individual buyer. Classic LTV modelling does not apply, and pretending it does leads people to overspend on acquisition.

What I use instead is repeat purchase rate from Amazon’s Brand Analytics repeat purchase behaviour data, measured per SKU over a rolling twelve weeks: what share of orders came from customers who had bought that ASIN before. For consumables it is the most useful number I have. For a one time purchase item it should be near zero, and if it is not, that usually means a sizing or quantity problem. People are re-buying because the first order was wrong.

I use it to decide where advertising money goes. A SKU with genuine repeat behaviour justifies bidding above break even on the first purchase, because the second and third orders arrive organically without ad cost. A SKU with no repeat behaviour has to pay for itself on order one, full stop. Two products with identical margins get completely different bid ceilings on that basis.

The other lever is Subscribe and Save enrolment rate, which is the closest thing Amazon gives you to a retention contract. If a consumable SKU has low enrolment, that is a listing and discount structure problem worth fixing before you touch bids.