Unit Economics Nobody Talks About

This post contains affiliate links. If you purchase through them, I may earn a small commission at no extra cost to you.

Your CAC/LTV ratio is 3:1. Your investor deck says you're crushing it.

You're not.

I know because I've been there. I had a SaaS product where the spreadsheet said we were printing money. CAC of $120, LTV of $480. Textbook healthy. Except we were hemorrhaging cash every month and couldn't figure out why.

Took me four months to find the answer: the inputs were wrong. Not the math—the assumptions feeding the math.

Unit economics are the most important numbers in your business. They're also the most commonly miscalculated. Here's what nobody tells you.

The Hidden Costs Inside CAC

Customer Acquisition Cost is supposed to capture everything you spend to acquire a customer. Most founders calculate it like this:

(Marketing Spend + Sales Salaries) ÷ New Customers = CAC

That's incomplete. Dangerously so.

Here's what gets left out:

• Engineering time on growth features — That referral system your engineers spent three weeks building? That's acquisition cost. The A/B testing framework? Acquisition cost. Every hour of engineering devoted to getting new users is CAC, not product development.
• Onboarding labor — If your customer success team spends the first two weeks hand-holding every new signup through setup, that's not retention spend. It's the tail end of acquisition. The customer isn't truly "acquired" until they're getting value independently.
• Free trial infrastructure — Server costs, support tickets, onboarding emails, demo calls—all consumed by users who never convert. You're paying full acquisition costs for zero revenue.
• Content production — Blog posts, videos, podcasts, social media. If a team of three produces content full-time, their loaded salaries belong in CAC.
• Tool subscriptions — Your CRM, email platform, analytics stack, ad manager accounts. These aren't "overhead." They're acquisition infrastructure.

When I recalculated our CAC with all of this included, it jumped from $120 to $210. Almost double.

That 3:1 ratio? It was actually 2.3:1. And once you factor in the LTV problems I'll get to next, it was closer to 1.4:1.

We were losing money on every customer.

LTV Is a Lie You Tell Yourself

Lifetime Value sounds precise. It's not. It's a projection based on assumptions that are almost always too optimistic.

The standard formula: ARPU × Gross Margin ÷ Monthly Churn Rate

Let's say you charge $49/month, your gross margin is 80%, and your monthly churn is 5%.

LTV = $49 × 0.80 ÷ 0.05 = $784.

Looks great. But here's where it falls apart:

Problem 1: Churn Isn't Constant

Your 5% churn rate is an average. But churn isn't evenly distributed.

Most SaaS products see heavy churn in months 1-3, then it tapers off. The customers who survive the first quarter tend to stick around. So your "average" churn blends sticky long-term customers with flighty short-term ones.

If your month-1 churn is actually 15% and your month-12+ churn is 2%, that 5% average is meaningless. You're losing most of your revenue in the first 90 days.

The LTV formula assumes every customer has the same expected lifetime. They don't. And blending cohorts masks the reality.

Problem 2: Gross Margin Isn't What You Think

Most founders calculate gross margin as revenue minus hosting costs. That's COGS in the narrowest sense.

Real gross margin includes:

• Customer support — Every ticket costs money. More customers = more tickets.
• Infrastructure scaling — Your $200/month server bill doesn't stay $200 when you 10x users.
• Payment processing — 2.9% + $0.30 per transaction. On a $49 charge, that's $1.72. Per customer. Per month. Over a year, that's $20.64 you never see.
• Compliance and security — SOC 2, GDPR, penetration testing. These scale with customer count.

Your 80% gross margin is probably closer to 65% when you count everything.

Revised LTV: $49 × 0.65 ÷ 0.05 = $637. That's a 19% haircut. And we haven't even fixed the churn assumption yet.

Problem 3: Expansion Revenue Is Hypothetical

Many founders juice their LTV by adding projected expansion revenue. "Customers will upgrade! They'll add seats! They'll buy add-ons!"

Maybe. But until you have data proving a consistent upgrade rate, expansion revenue is speculation. And baking speculation into your unit economics is how you end up making decisions based on fantasy.

I watched a competitor build their entire growth model on projected expansion revenue. They assumed 40% of customers would upgrade to the enterprise tier within 12 months. The actual number was 6%.

Their LTV was off by 3x. They burned through $2M before they figured it out.

The Metric That Actually Matters: Payback Period

Forget LTV/CAC ratio. The single most important unit economic for an early-stage company is payback period: how many months until you recover the cost of acquiring a customer.

Why? Because payback period tells you about cash flow, not theoretical lifetime value.

If your CAC is $210 and your monthly gross profit per customer is $32, your payback period is 6.6 months.

That means for six and a half months, every new customer is a cash drain. You're spending money today and hoping to earn it back later.

If you're growing 20% month-over-month with a 6-month payback, you need enough capital to fund six months of acquisition costs for an exponentially growing customer base. Most startups don't have that.

Payback period under 6 months? You can probably self-fund growth.

Payback period 6-12 months? You need outside capital or very controlled growth.

Payback period over 12 months? You're gambling that churn won't eat your lunch before you break even.

Cohort Analysis or You're Flying Blind

Blended metrics hide everything. If you're not doing cohort analysis, you're making decisions with a blindfold on.

A cohort is a group of customers who signed up in the same period—usually the same month.

Track each cohort separately:

• Retention curve — What percentage of the January cohort is still active in February? March? June?
• Revenue per cohort — Is each cohort generating more or less revenue over time?
• CAC per cohort — Are you spending more or less to acquire customers as you scale?

When I started doing this, I discovered something alarming: our January cohort had 85% month-3 retention. Our April cohort had 62%.

Same product. Same pricing. Same onboarding. But the later cohort was churning 23 percentage points faster.

Why? Because our early customers were friends, referrals, and industry contacts. They were pre-sold. Our later customers came from paid ads. They were colder. Less committed. More likely to churn.

If I'd been looking at blended churn, I would have missed this entirely. The early cohorts were masking the decay of later ones.

The Dangerous Moment: When CAC Starts Climbing

Early on, CAC is low. You're picking the low-hanging fruit. Friends. Network connections. Organic traffic from launch coverage.

Then you exhaust those channels. You turn to paid acquisition. And CAC starts climbing.

Here's what happens next:

1. You increase ad spend to maintain growth rate.
2. CAC rises because you've saturated the cheapest audiences.
3. You increase spend again to hit targets.
4. CAC rises further.
5. Your LTV/CAC ratio deteriorates, but you don't notice because you're looking at blended numbers.

By the time you realize CAC is unsustainable, you've already committed to a growth rate that requires it.

This is the death spiral. And it starts with not tracking CAC by channel and by cohort.

Some channels have a $30 CAC. Others have a $300 CAC. If you blend them, you see $120 and think everything is fine. But you're subsidizing an unprofitable channel with a profitable one—and as the unprofitable channel scales, it eats the margin.

What Good Unit Economics Actually Look Like

Forget the 3:1 rule of thumb. It's too simplistic. Here's what I look for:

1. Fully loaded CAC — Including every cost described above. If it's above your annual contract value, you have a problem.
2. Payback period under 12 months — Ideally under 6. This determines your cash efficiency.
3. Cohort retention above 80% at month 6 — If less than 80% of a cohort is still paying after six months, your product or positioning has a problem.
4. Negative revenue churn — Expansion revenue from existing customers exceeds lost revenue from churned ones. This is the holy grail.
5. Channel-specific CAC tracking — Every acquisition channel measured independently. Kill the ones that don't work.

These are harder to calculate than a simple LTV/CAC ratio. They take more discipline to track. But they tell you the truth about whether your business is viable.

The Conversation Nobody Wants to Have

Here's the uncomfortable part: most startups have bad unit economics and don't know it.

They use optimistic churn assumptions. They exclude real costs from CAC. They project expansion revenue that never materializes. They look at blended metrics that hide deterioration.

And they keep raising money to cover the gap.

That works until it doesn't. Eventually, someone—an investor, a board member, a new CFO—looks at the real numbers. And the conversation changes from "how fast can we grow" to "are we actually making money."

Better to have that conversation with yourself first.

Sit down with your numbers. Calculate fully loaded CAC. Run cohort analysis. Track payback period. Look at channel-specific performance.

If the numbers are ugly, good. Now you know. And knowing lets you fix it before someone else finds out.

What I'd Do Differently

If I could restart my first SaaS, I'd change three things:

1. Track cohorts from day one. Not blended averages. Individual cohorts. Monthly. Without exception.
2. Calculate fully loaded CAC from day one. Every cost. No shortcuts. Even when the numbers are small and it feels like overkill. The habits you build early compound.
3. Set a payback period ceiling. If acquiring a customer takes more than 8 months to pay back, don't acquire them through that channel. Period.

Unit economics aren't something you figure out later. They're the foundation you build on from the start.

Get them right, and scaling is a math problem.

Get them wrong, and scaling is a death sentence.

Related Reading

If this resonated, you might also find value in:

• Revenue Is Validation—Everything Else Is Noise – Why the only metric that matters early on is whether people pay you
• When Free Users Become Freeloaders – The economics of freemium gone wrong
• SaaS Unit Economics Books on Amazon – Deep dives on building financially sound businesses
• Startup Metrics Resources on Amazon – Frameworks for measuring what matters