Ecommerce Customer Lifetime Value: How to Calculate and Grow LTV

Customer acquisition costs in ecommerce have risen more than 60% over the past five years, driven by iOS privacy changes, rising competition on paid channels, and the emergence of well-funded discount players that have raised CPMs across Meta and Google. According to analysis by Eightx, a fractional CFO firm that works with ecommerce brands, the average ecommerce brand now loses around $29 per new customer acquired after factoring in marketing costs and typical return rates — a figure to treat as directional rather than precise, but the direction is clear.

In that environment, LTV is not a vanity metric. It’s the mechanism by which sustainable ecommerce businesses justify their acquisition spend, allocate marketing budgets by channel, and build a case for profitability that isn’t entirely dependent on finding cheaper traffic. This guide walks through how to calculate LTV correctly, how to benchmark it, what the levers are for improving it, and — folding in the article’s secondary keyword — how B2B lead scoring connects to LTV as the acquisition-side complement.

A note on terminology: LTV and CLV (customer lifetime value) are used interchangeably throughout the industry. Some sources distinguish them — CLV for an individual customer, LTV for the average across a cohort — but there’s no standard industry definition. We’ll use LTV throughout.

Why Most LTV Calculations Are Wrong

Before the formula, a more important point: most ecommerce LTV numbers are wrong in a specific direction — they overstate value. The two most common errors are using revenue instead of margin, and not accounting for returns.

Revenue vs. gross margin. If a customer spends £500 with you over their lifetime, their revenue-based LTV is £500. But if your gross margin is 35%, the margin-adjusted LTV is £175. These two numbers have completely different implications for how much you can afford to spend acquiring that customer. A business running a 35% gross margin that calculates LTV on revenue will set its acquisition budget 2.86x too high — and wonder why it’s never profitable.

Returns and refunds. In apparel and footwear, return rates run at 25–40% of orders. If your LTV calculation doesn’t net out returns, you’re inflating both average order value and purchase frequency. A customer who places six orders but returns half of them is not worth twice what the raw order count suggests.

Acquisition channel conflation. An LTV calculated across all customers blends high-LTV customers acquired through organic channels or referrals with low-LTV customers acquired through paid discount campaigns. The blended number hides the channel-level reality: customers acquired with 30%+ off promotions often have materially lower LTV than customers acquired at full price, because the discount signals price sensitivity that persists through the relationship.

Fix these three before you trust your LTV number. The correct input for every formula that follows is gross profit per order, not revenue.

The LTV Formula: Three Versions, in Order of Complexity

There is no single universal LTV formula. The right version depends on your business model, your data availability, and what decision you’re trying to make. Here are the three standard versions, from simplest to most rigorous.

Version 1: The simple formula

Version 2: The gross-margin formula (more precise)

Some practitioners separate the gross margin adjustment and express it explicitly in the formula, which makes it easier to model the impact of margin changes on LTV:

Version 3: The cohort-based formula (most actionable)

The simple formula uses averages — average AOV, average frequency, average lifespan. Averages hide the distribution. A business where 20% of customers account for 80% of LTV (a common Pareto pattern) looks very different in a cohort breakdown than it does in the blended average. Cohort-based LTV calculates the cumulative gross profit of a defined group of customers — acquired in the same month, from the same channel, or in the same campaign — tracked over a defined period (typically 12, 24, and 36 months).

LTV:CAC Ratio — What Yours Should Be

LTV in isolation is not a useful number. A £300 LTV means something very different if you spent £50 acquiring that customer versus £280. The ratio that matters is LTV:CAC — lifetime value divided by customer acquisition cost.

The universally cited benchmark is 3:1. First Page Sage, which has compiled client data across 29 industries from 2020 to 2024 (80+ clients), places this as the standard healthy ratio for ecommerce. Bloomreach, Drip, and several fractional CFO firms converge on the same number. Here’s how to interpret different ratio levels:

LTV:CAC benchmarks by ecommerce vertical

The 3:1 aggregate benchmark masks significant variation by category. The following ranges are compiled from analytics platforms and fractional CFO firms that publish vertical benchmarks (LTV.ai, Eightx) — treat them as directional starting points rather than peer-reviewed figures, and calibrate against your own cohort data as it builds.

How to Calculate Your Customer Lifespan (Without Guessing)

The most frequently guessed input in LTV calculations is customer lifespan. Founders often use 3 or 5 years because those numbers feel reasonable, not because the data supports them. Guessing lifespan inflates LTV systematically and produces a false sense of unit-economic health.

The correct approach is to measure customer lifespan from your order data. The standard method for a non-subscription ecommerce business:

1. Define “active”: A customer is active if they placed an order within a defined window — typically 12 months for high-frequency categories, 24 months for lower-frequency ones. This threshold should be informed by your median inter-purchase gap.
2. Identify your churn point: Look at the distribution of time gaps between orders. The point at which the probability of a repeat order drops below a defined threshold (often 10–15%) is your effective churn point. For many ecommerce businesses this is 6–9 months of inactivity.
3. Calculate average lifespan from cohort data: For customers who have passed their churn point, measure the time from first order to last order. Average this across cohorts. Do not include customers who are still active — their lifespan is not yet known.
4. Segment by acquisition channel: Organic-acquired customers consistently have longer lifespans than paid-discount-acquired customers. A blended average hides the channel-level dynamics that should inform your acquisition strategy.

The Four Levers That Actually Move LTV

LTV = AOV × purchase frequency × customer lifespan. Everything that moves LTV connects to one of these three variables, plus gross margin. The levers are not equally accessible at all stages of a business, and they have different time horizons before they pay back.

Lever 1: Drive the second order

The most undervalued LTV lever in ecommerce is the one immediately after the first purchase. Customers who make a second order have dramatically higher predicted LTV than customers who have made only one. The reason is simple: a second order signals intent to continue the relationship. A single order could be opportunistic (a discount, a gift, an impulse). Two orders suggest fit.

The tactical implication is that your most important retention email is not your 12-month win-back campaign — it’s the post-purchase flow designed to generate a second order within 60–90 days of the first. Personalised product recommendations based on the first order, education content, and a modest incentive timed 3–4 weeks post-purchase consistently outperform generic promotional sends in second-order conversion rate.

Businesses that track “30-day repeat purchase rate” as a leading indicator — rather than trailing LTV — catch LTV deterioration before it becomes visible in revenue. A drop in 30-day repeat rate across consecutive quarterly cohorts signals an LTV problem that will become a revenue problem six to twelve months later.

Lever 2: Raise average order value through structure, not discounts

AOV is the most immediately actionable lever but also the one most commonly addressed through discounting — which is counterproductive. Discounts raise AOV on a single order but attract price-sensitive customers with lower subsequent LTV. The structural approaches that raise AOV without degrading the customer profile:

• Free shipping thresholds: Setting the free shipping threshold 20–30% above current average order value creates an uplift mechanism without requiring a discount. Customers who would have spent £42 spend £55 to hit the £50 threshold.
• Bundles based on complementary purchase data: Products frequently purchased together (identified from order history) bundled at a marginal saving outperform random cross-sells. The bundle logic should come from the data, not from merchandising intuition.
• Post-purchase upsells: The conversion rate on a post-purchase upsell — offered after the order is confirmed but before the customer leaves — is typically higher than pre-purchase cross-sell because the purchase decision has already been made. Low-friction formats (one-click add) work best.

Lever 3: Increase purchase frequency through relevance

Purchase frequency is driven by the perceived relevance of communication and the strength of habit formation. Generic broadcast emails generate diminishing engagement over time. The behaviours that sustain frequency:

• Replenishment triggers: If your products are consumable, model the expected replenishment date from purchase quantity and average usage, and trigger a reminder email 5–7 days before that date. This is purely data-engineering work — the relevant signal is already in your order history.
• Category cross-pollination: Customers who buy across multiple product categories have materially higher LTV than single-category buyers. Identify customers who have bought from one category and expose them to adjacent ones through personalised recommendations.
• Loyalty mechanics that reward behaviour, not just spend: Points-per-pound schemes are easily ignored. Loyalty mechanics that unlock tangible, non-discount benefits — early access, exclusive products, free returns — create structural switching costs.

Lever 4: Extend lifespan through win-back before churn

Most win-back campaigns are reactive — triggered after a customer has already lapsed. Predictive churn models can identify customers showing early lapse signals (declining open rates, longer inter-purchase gaps, reduced engagement with personalised recommendations) and trigger intervention before the lapse becomes permanent. This is the territory where machine learning genuinely pays for itself: the churn signal space is too large to monitor manually at scale, and the value of recovering a customer before they lapse is higher than recovering them after.

For businesses without ML capacity, a simple proxy: any customer who bought at least twice and has not ordered in 5 months is approaching your churn horizon. A targeted sequence with a genuine reason to return (new products relevant to their history, a time-limited free-shipping offer, a personalised replenishment reminder) at the 5-month mark costs almost nothing to operate and recovers a meaningful percentage of at-risk customers.

B2B Lead Scoring: The Acquisition-Side Complement to LTV

This section addresses the article’s secondary keyword — B2B lead scoring for ecommerce — and the link it has to LTV. For ecommerce businesses with a B2B customer segment (wholesale buyers, resellers, corporate gifting clients, white-label partners), lead scoring is the mechanism by which you replicate, on the acquisition side, the same discipline that LTV analysis applies on the retention side.

The connection is direct: if your LTV analysis reveals that B2B wholesale customers have a 5:1 LTV:CAC ratio while B2C retail customers average 2.8:1, your lead scoring system should be calibrated to preferentially route sales and marketing resources toward companies that fit the wholesale customer profile. Without that calibration, a generic lead scoring system will optimise for conversion rate rather than for downstream LTV — and conversion rate is a poor proxy for LTV.

What B2B lead scoring actually measures

B2B lead scoring assigns numerical values to leads based on two categories of signals:

Explicit (firmographic) signals: Company size, industry, annual revenue, location, tech stack, job title of the contact. These are the inputs that determine whether a prospect matches your ideal customer profile — the type of business that, based on your actual customer data, tends to become a high-LTV customer.

Implicit (behavioural) signals: Pages visited, content downloaded, email open and click rates, webinar attendance, pricing page views, repeated visits over a short window. These signals indicate purchase intent and engagement level — a prospect who has viewed your pricing page three times and downloaded two case studies is warmer than one who opened a single email.

The standard framework for structured lead qualification is BANT — Budget, Authority, Need, Timeline — originally developed by IBM. BANT is fast and easy to learn, and it remains useful as an initial screening tool for high-volume, shorter-cycle B2B sales. Its limitations are well-documented: it assumes budget is allocated before the conversation starts, which is not always true in modern B2B purchases, and it focuses on a single decision-maker in environments where the average B2B purchase now involves multiple stakeholders.

For most ecommerce businesses with a B2B segment, a hybrid approach works well: BANT for initial screening to filter obvious mismatches, then a deeper qualification layer (closer to CHAMP) for opportunities that pass the initial gate. The key is that your qualification criteria — whatever framework you use — should be calibrated against actual customer data. The questions to answer from your existing B2B customer records: Which firmographic characteristics predict high LTV? At what engagement score do leads tend to convert? Which acquisition channels produce B2B customers with the best retention?

A practical B2B lead scoring model for ecommerce

The following is a worked example of a lead scoring model calibrated for an ecommerce business selling wholesale to independent retailers. Adjust thresholds based on your actual customer data — this is a starting structure, not a prescription.

Calibrating lead scoring against LTV outcomes

A lead scoring model that optimises for conversion rate alone will over-index on leads that are easy to close — which are not always the same as leads that become high-LTV customers. A lead who converts quickly on a steep discount and then churns after two orders is a worse outcome than a lead who required eight weeks of nurturing and then placed consistent repeat orders for three years.

The calibration step: once you have 12+ months of data from your lead scoring system, match closed leads to their actual LTV outcomes. Identify which firmographic and behavioural signals in your scoring model correlate with high LTV at 12 months, and which correlate with high conversion-but-low-retention. Adjust point weights accordingly. This requires clean integration between your CRM, your marketing automation platform, and your ecommerce order data — exactly the kind of data engineering problem that compounds in value over time as the model improves.

For businesses with a substantial B2B revenue stream, this calibration loop is one of the highest-leverage investments available. A scoring model tied to real LTV outcomes — not just conversion rate — ensures your sales team targets prospects most likely to become long-term, high-margin accounts. The infrastructure required: a CRM that records lead score at conversion, an order management system that attributes revenue to the original lead record, and a reporting layer joining the two. Without those connections, calibration is impossible regardless of how well-designed the scoring model is. See our B2B growth work for how we approach this integration in practice.

When LTV Analysis Tells You Something Uncomfortable

LTV analysis done honestly will sometimes produce results that are difficult to act on. The most common uncomfortable conclusions:

Limitations of This Article

Several figures in this guide carry sourcing caveats:

• The LTV:CAC 3:1 benchmark converges across multiple sources; First Page Sage provides the most methodology-disclosed version (80+ clients, 2020–2024, noted skew toward midsized and B2B firms). Treat as industry consensus, not a regulatory or certified standard.
• Vertical LTV:CAC ranges (pet care 3.5–4:1, supplements 3–5:1, etc.) are compiled from LTV.ai and Eightx analytics platforms with their own client bases — directional guidance, not peer-reviewed research.
• The “CAC up 60%+ over 5 years” and “$29 per-customer loss” figures are from industry-practitioner sources (Eightx, LTV.ai); not from independent academic or market research. The directional claim is broadly supported by documented CPM inflation on Meta and Google, but specific numbers should be treated as illustrative.
• BANT “59% conversion rate increase” claim (cited by Salesmotion citing InsideSales, now Xant) — if using in internal materials, locate the original InsideSales research report to verify the methodology.