Customer Lifetime Value (CLV or LTV) is the single most important metric in eCommerce. It tells you how much revenue a customer will generate over their entire relationship with your brand — which determines how much you can afford to spend acquiring them, which segments to focus retention efforts on, and ultimately whether your business model works.
Despite its importance, most brands either skip it entirely or calculate it wrong. This guide covers every approach, from the simple back-of-napkin formula to the probabilistic models used by growth teams at scale.
What Is Customer Lifetime Value?
CLV (also written LTV, CLTV, or pCLV for the predicted version) is the total net revenue you expect to earn from a customer over the entire time they remain active. A customer who buys once for €50 has a much lower CLV than one who buys five times per year for three years at €80 each — even if their first order values were identical.
Understanding CLV shifts how you think about growth. Instead of asking "how do I get more orders this month?", you start asking "how do I acquire and retain the customers who are worth the most over 12 or 24 months?"
Key insight: If your average CLV is €240 and you're paying €80 to acquire a customer, your LTV:CAC ratio is 3:1 — the healthy minimum. If CLV drops to €120, you're on the edge of unprofitability without changing your ad spend at all.
The Simple CLV Formula
The simplest way to calculate CLV uses three variables you almost certainly already have:
Let's work through a real example. Suppose you run a Shopify skincare brand:
- Average Order Value (AOV): €62
- Purchase Frequency: 3.2 orders per year
- Average Customer Lifespan: 2.4 years
CLV = €62 × 3.2 × 2.4 = €476
Now you know that each new customer is worth €476 on average over their lifetime. If you're running Google Ads at €90 CPA, you have a healthy 5.3:1 LTV:CAC ratio. If you're at €200 CPA, it's still a profitable 2.4:1 — but you'd want to monitor it closely.
How to find each variable
- AOV: Total revenue ÷ number of orders in a period. Most eCommerce platforms show this directly.
- Purchase frequency: Total orders ÷ unique customers in a period. Harder to get from dashboards — you often need to export transaction data.
- Customer lifespan: The trickiest one. One approach: 1 ÷ churn rate (if 20% of customers don't return in year 2, lifespan ≈ 5 years). Another: calculate the median time from first to last purchase in your historical data.
Limitations of the Simple Formula
The simple formula is a useful starting point but breaks down in a few important ways:
- It treats all customers the same. In reality, the top 20% of customers typically generate 60–80% of revenue. Averaging hides this.
- It's backward-looking. It uses historical averages, not predictions about individual customer future behavior.
- It doesn't handle irregular purchase patterns well. A customer who bought twice in the past week looks the same as one who bought once per year for two years — but they behave very differently.
- You can't act on it at the customer level. You get one number for everyone, not individual scores you can use for segmentation or personalization.
The right tool for the job: Use the simple formula for board presentations and quick sanity checks. Use probabilistic models (below) for segmentation, retention prioritization, and CAC decisions by channel.
Probabilistic CLV: The BG/NBD Model
The gold standard for predicting individual customer lifetime value is the BG/NBD model (Beta-Geometric / Negative Binomial Distribution), developed by Fader, Hardie, and Lee at Wharton. It's the same framework used by Amazon, Spotify, and most growth-stage DTC brands.
Instead of averaging historical behavior, BG/NBD models two simultaneous processes for each customer:
- While active: How often does this customer purchase? (Negative Binomial Distribution)
- When do they churn? At any point after a purchase, a customer might quietly stop buying. (Beta-Geometric)
The model fits these distributions to your transaction history and outputs, for each customer, a probability of being still alive and an expected number of future purchases in any time window you specify.
What you need to run BG/NBD
The model requires transaction data in a specific format — one row per customer with:
- Recency: Time between first and last purchase
- Frequency: Number of repeat purchases (purchases after the first)
- T: Time from first purchase to the observation date (how long we've been watching them)
Combined with the Gamma-Gamma model (which predicts average order value per customer), BG/NBD gives you a predicted CLV for each individual customer — ranked, segmented, and actionable.
| Approach | What you get | Best for | Data needed |
|---|---|---|---|
| Simple formula | One aggregate CLV number | Reporting, LTV:CAC checks | AOV + purchase frequency |
| Cohort analysis | CLV by acquisition month/channel | Measuring retention trends | Transaction data with dates |
| BG/NBD + Gamma-Gamma | Individual predicted CLV + segments | Retention, segmentation, ad bidding | Full transaction history (CSV) |
CLV Benchmarks by Industry
CLV varies enormously by business model and product category. Here are typical ranges for eCommerce:
| Industry | Avg. CLV (12-month) | Avg. CLV (3-year) | Key driver |
|---|---|---|---|
| Fashion (DTC) | €120–€250 | €280–€600 | Repeat rate, AOV |
| Beauty & Skincare | €180–€380 | €500–€1,200 | High repurchase frequency |
| Accessories / Jewellery | €90–€200 | €200–€480 | Low frequency, high AOV |
| Home & Lifestyle | €100–€280 | €240–€700 | Seasonal purchase patterns |
| Food & Supplements | €200–€500 | €600–€1,800 | Subscription or high frequency |
These are rough ranges — your actual numbers depend heavily on acquisition channel, price point, and product category. The only reliable benchmark is your own historical data.
How to Improve Your CLV
CLV is a function of three levers: how much customers spend per order (AOV), how often they buy (frequency), and how long they stay (retention). Each lever requires different strategies:
Increase AOV
- Bundle products with a small discount (e.g., "Buy 2, save 15%")
- Free shipping thresholds slightly above your current AOV
- Post-purchase upsell immediately after checkout
Increase purchase frequency
- Replenishment reminders for consumable products (email at predicted reorder time)
- Loyalty programs that reward frequency, not just spend
- Cross-category campaigns: customers who bought X also love Y
Improve retention
- Win-back campaigns targeting customers who haven't bought in 60–90 days
- VIP treatment for your top LTV segment (early access, exclusive products)
- Post-purchase experience: packaging, thank-you emails, onboarding sequences
Highest leverage: Improving retention by 5% typically increases CLV by 25–95% (Bain & Company). Focus there before trying to squeeze more AOV.
Using CLV for Ad Bidding
Once you have individual predicted CLV scores, you can feed them into Google Ads and Meta as customer signals. This lets the algorithm optimize toward high-CLV customers rather than just next-purchase probability.
The practical flow:
- Calculate predicted 12-month CLV per customer (using BG/NBD)
- Segment customers into value tiers (Champions, Loyal, At-Risk, Lost)
- Upload top-tier customers to Google Ads Customer Match
- Create lookalike audiences based on your Champions segment
- Bid more aggressively for these lookalikes vs. cold audiences
This approach is sometimes called "value-based bidding" and is one of the highest-ROI uses of CLV data available to eCommerce marketers.
Frequently Asked Questions
What's the difference between CLV and LTV?
They're the same metric. CLV (Customer Lifetime Value) and LTV (Lifetime Value) are used interchangeably. Some companies use CLTV (Customer Lifetime Value) or pCLV (predicted CLV) to be more specific.
How much data do I need to calculate CLV?
For the simple formula, you only need a few months of orders. For the BG/NBD probabilistic model, you need at least 6–12 months of transaction history and ideally 500+ customers who have made at least one repeat purchase. The more data, the more accurate the predictions.
Should I use revenue or profit for CLV?
Revenue CLV is easier to calculate and is fine for comparing customers relative to each other (segmentation). For LTV:CAC decisions, use gross profit CLV (revenue × gross margin) to know what you can actually afford to spend on acquisition.
How often should I recalculate CLV?
Monthly is a good cadence for most eCommerce brands. If you're running active retention campaigns, you'll want to track how they shift your CLV distribution over time.