Calculate Customer Lifetime Value: Models for SaaS, E-Commerce, and Services

Half the businesses I work with in Kuwait are spending acquisition money on channels they think are profitable but actually aren't. The reason is almost always the same: their LTV is a fiction.

Here's how it usually works: a founder decides they want to acquire customers for 500 KWD or less. They back-calculate what LTV needs to be to justify that spend, assume they'll achieve it, and build their entire marketing strategy on that number. Then reality hits. Churn is higher than expected. Repeat purchase rates drop. The "model" collapses.

I've watched this exact mistake kill projects that were otherwise well-funded. And I've learned that most businesses don't fail because their LTV is low—they fail because they don't actually know what it is.

The good news: calculating real LTV isn't complicated, but it does require honesty. And the formula changes depending on your business model. A SaaS company's LTV tells a completely different story than an e-commerce shop's, which tells yet another story for an agency or consulting firm. You need to know which one applies to you, how to measure it, and how to avoid the traps that make it worthless.

Let me walk you through each.

Why LTV actually matters—and it's not what you think

Most people talk about LTV as if it's a single number: "Our LTV is 5,000 KWD." But LTV isn't really a target or a badge of honor. It's a constraint.

Here's what I mean. If your customer acquisition cost (CAC) is 1,000 KWD and your LTV is 3,000 KWD, you have a 3-to-1 ratio. That sounds profitable in isolation. But it's not. Why? Because you don't see all of that 3,000 KWD at once. You see it spread across months or years. If it takes 18 months to recoup your acquisition spend, you're tying up cash, carrying opportunity cost, and betting that the customer doesn't churn in the meantime.

That's where payback period comes in—the time it takes to earn back what you spent to acquire a customer. A business with a 6-month payback and a 2.5:1 LTV-to-CAC ratio is in a completely different position than one with a 18-month payback and a 4:1 ratio, even though the second looks better on paper.

When a client comes to us asking about scaling acquisition, the first question I ask isn't "What's your LTV?" It's "How long does it take to pay back CAC?" The answer tells me whether they can actually afford to grow.

LTV also shapes your retention strategy. If you know your LTV is 2,000 KWD and a customer would stay 18 months on average, you know you can spend up to some portion of that to prevent churn. Many businesses in the region don't make this connection—they treat retention as a nice-to-have instead of a profit center. That's the second mistake.

The three components every LTV model shares

No matter what business model you're in, LTV is built from the same building blocks. The formulas change, but these three elements always matter:

1. Revenue per customer. For SaaS, that's monthly recurring revenue (MRR) per user. For e-commerce, it's average order value (AOV). For services, it's the contract value. This is the easiest to measure and often the part businesses get right.

2. How often they generate that revenue. For SaaS, it's implicit—monthly is built into "MRR." For e-commerce, it's the number of repeat purchases in a year. For services, it's how many engagements you can chain together. This is where a lot of businesses start guessing.

3. How long they stay your customer. Retention is the silent killer of LTV models. Miss this, and your entire calculation is wrong. SaaS measures this by churn rate (the percentage who cancel each month). E-commerce measures repeat customer rate and purchase frequency. Services measure contract renewal and repeat engagement. The businesses that get this right separate customers by cohort—everyone acquired in January, everyone in February—and track them individually. That's because churn changes over time and varies by acquisition channel.

Here's the uncomfortable truth: most businesses don't track cohort retention. They track average retention across all customers, which hides the fact that recent cohorts churn faster because they're still testing whether to stay. That's a recipe for overestimating LTV.

SaaS customer lifetime value: the recurring revenue model

SaaS is the simplest model to calculate because revenue is predictable and recurring. Every month, the customer either stays or churns. That creates a mathematically clean formula.

The standard SaaS LTV formula is:

LTV = ARPU / Monthly Churn Rate

Where ARPU is average revenue per user. Or, if you prefer a longer form that explicitly shows the time component:

LTV = (ARPU × Gross Margin) × (1 / (1 - Retention Rate))

Let's use a real example. Say you're a SaaS platform serving accounting firms across the GCC with a basic plan at 500 KWD/month. Your ARPU (accounting for different plan tiers) is actually 650 KWD/month. Your gross margin is 75% (after hosting, support, payment processing). Your monthly churn rate is 6%.

LTV = (650 × 0.75) × (1 / (1 - 0.94)) = 487.5 × 16.67 = 8,125 KWD

That looks great until you ask: what's your payback period? If your CAC is 2,000 KWD and you're acquiring customers at the blended rate across channels, and those customers have an average lifetime of 16.7 months (the inverse of 6% monthly churn), then payback happens at month 3.2. That's healthy. You can afford to spend on growth.

But—and this is critical—that math only works if 6% churn is stable. If churn is 6% for your first cohort but 10% for your most recent cohort (because you just changed your pricing model or a competitor launched), your real LTV for new customers is only 4,875 KWD. Suddenly your acquisition strategy needs to change.

For SaaS, track three numbers obsessively: ARPU (break it down by plan tier and customer segment), monthly churn (by cohort, not average), and gross margin. Everything else is just arithmetic. The mistake most SaaS founders make is thinking LTV is constant. It's not. It's a function of churn, and churn drifts.

E-commerce LTV: the repeat purchase puzzle

E-commerce is harder because most transactions are one-off, but some customers come back. Your LTV depends entirely on how many do and how often.

The basic formula is:

LTV = AOV × Purchase Frequency × Customer Lifespan

Let's say you run an e-commerce store selling premium kitchen equipment to households across Kuwait. Your AOV is 350 KWD. Your data shows that 35% of customers make a second purchase within 12 months, and about 12% of those make a third. Your repeat purchase frequency is 1.35 purchases per customer in the first year.

Now, how long does this last? Unlike SaaS (where customers are "active" until they cancel), e-commerce has to define what "active" means. Usually it's: customer makes a purchase within the last 12 months. So a customer acquired today might generate revenue for 2–3 years on average before going silent.

LTV = 350 × 1.35 × 2.5 years = 1,181 KWD

If your CAC is 300 KWD per customer (paid ads + organic), you're at a 3.9:1 ratio, which sounds excellent. But again—payback matters. A lot of that 1,181 KWD comes in year 2 and 3. Your cash flow in months 1–4 is much tighter.

Here's what makes e-commerce LTV modeling tricky: repeat purchase behavior varies wildly by product category and customer segment. A customer who buys luxury items once every three years has an LTV that's front-loaded and long-tailed. A customer buying consumables monthly has consistent revenue. You need to build separate LTV models for each segment because they have completely different growth implications.

I've seen e-commerce businesses treat all customers as if they have the same LTV, then get confused when they scale acquisition and repeat purchase rates collapse. Turns out they were acquiring the wrong customer segment. The ones with high LTV were always a minority.

Service businesses: when LTV is a chain of projects

My own domain. Service LTV is conceptually similar to SaaS recurring revenue, but it works in discrete projects instead of continuous subscription.

A consulting firm, design studio, or development shop gets a project—let's say a website redesign at 15,000 KWD. That revenue hits when the project completes. The real LTV question is: how many more projects does that client give you?

The formula is:

LTV = Project Value × Repeat Rate × Average Number of Engagements

Let's say your average project is worth 10,000 KWD. Your data (tracked over two years) shows that 40% of clients come back for a second project, and 15% of those come back again. Your repeat rate is 0.40 + (0.40 × 0.15) = 0.46 engagements per client on average.

LTV = 10,000 × 0.46 × 1 = 4,600 KWD (if we assume engagements stop after two)

Or, more precisely:

LTV = 10,000 × (1 / (1 - 0.46)) = 18,518 KWD (if the renewal rate is stable)

The payback here is fast—you get most of the CAC back in month one when the project completes. But the next engagement might not come for 8 months. That's the cash flow risk with service businesses.

Here's what I've learned consulting on this: service businesses grossly overestimate repeat rates by not tracking cohorts. You think clients will keep coming back, so you estimate 50% renewal. In reality, you get 25% because half your clients churn to other vendors or go in-house. The only way to know your real repeat rate is to track it by year of acquisition and watch it stabilize.

Honestly, I haven't seen enough service businesses track their LTV properly to say definitively what the "right" repeat rate is across the region. But I know most are optimistic. Build conservatively and prove it with data.

Making your LTV model actionable

Once you have a number, the temptation is to declare victory and move on. Don't.

Real LTV work has four steps. First: calculate LTV by cohort and acquisition channel. A customer acquired via Google Ads might have 30% higher LTV than one from LinkedIn because the Google cohort retains better or spends more. That's crucial information. Second: calculate payback period for each channel. If Google has a 4-month payback and LinkedIn has a 10-month payback, Google funds growth. LinkedIn needs to improve before you pour money there. Third: project LTV forward quarterly. Churn drifts. Repeat rates shift. A model that was accurate in Q2 might be 20% off by Q4. Track it. Fourth: use LTV to set retention budgets. If LTV is 5,000 KWD and a customer is at risk of churn, how much can you spend to save them? The answer is: up to the difference between LTV and expected future value if they churn. Most businesses guess. Build the math.

The uncomfortable bit: doing this right means storing data for 12+ months before you have reliable numbers. You can't know churn until people have had time to churn. You can't know repeat rates until enough time passes. Most founders want answers now. The math doesn't work that way.

Where LTV modeling breaks down

LTV is an average. Averages lie.

Two businesses with identical LTV of 3,000 KWD can have completely different growth trajectories. One might have high-value customers that need heavy onboarding (longer payback, higher retention). The other might have low-value customers that are easy to acquire but churn fast. The blended LTV says nothing about those dynamics.

External factors matter too. A SaaS company's churn rate isn't random—it responds to pricing changes, feature launches, and competitor moves. An e-commerce store's repeat purchase rate changes with seasonality and marketing focus. You can model based on historical data, but you can't predict the future with precision. LTV models are guides, not prophecies.

I'd also say this: LTV is most useful for a business that has achieved product-market fit and stability. A very early-stage startup calculating LTV is usually wasting time. The business is too volatile. Track it, sure, but don't build your entire strategy around a number that will be wrong next quarter. Same for businesses in high-churn industries where nothing stabilizes. If you're a marketplace and churn is 30% monthly, your LTV formula is unstable by definition.

One more honest caveat: LTV doesn't account for operational cost of serving the customer. A customer with high LTV might require expensive support, which reduces actual profit. Real LTV should be LTV minus the blended cost of retention, support, and operations. Most businesses don't subtract that. They should.

The math in practice

Let me ground this with a specific scenario from a client project. A Kuwaiti e-commerce store selling home fitness equipment reached out. They were spending 500 KWD per customer on ads, converting at 2.5% on paid traffic. They thought their LTV was 2,500 KWD based on AOV (400 KWD) and an assumption of 6 purchases per customer lifetime. That gave them a 5:1 ratio, which felt safe.

We rebuilt it with actual data. AOV was correct at 400 KWD. But repeat purchase data told a different story. Of first-time customers, 18% bought again within 12 months. Of those, 8% bought a third time. That's 0.26 engagements per customer on average, not the assumed 6. Customer lifespan was 2.2 years on average (using 12-month activity window definition).

Real LTV = 400 × 0.26 × 2.2 = 228.8 KWD.

Not 2,500. Real payback was negative for the first five months. They were losing money on acquisition until customers made second purchases. And a lot never did.

The response wasn't to abandon acquisition. It was to reframe the strategy: invest in customer onboarding and email nurture to drive repeat purchases in months 2–4, when payback actually happens. Build LTV by improving the repeat purchase rate to 25%, not by acquiring more customers. That math was different. And it worked.

Key takeaways for your business

If you run a SaaS business, obsess over churn by cohort. That's your LTV dial. If you run e-commerce, build separate LTV models by product category and customer segment. The blended number will lie to you. If you run a service business, track repeat rates honestly and calculate payback period before you commit to scaling acquisition.

All three models share one rule: measure payback period before you measure LTV. Payback tells you whether you can afford to grow. LTV tells you how much you can ultimately afford to spend. One is a constraint, the other is a ceiling. Both matter.

Start collecting the data now. You won't have reliable LTV figures for 12 months. But that doesn't mean you wait. It means you track, measure, and adjust quarterly. After a year, you'll have a real number. Use it to decide what to scale and what to kill.