Marketing Attribution Is Broken. Here's What to Use Instead.

Quick Navigation

What Broke Attribution in 2021

iOS 14 changed everything. Apple blocked third-party cookie tracking. Google announced it was killing the browser cookie. Facebook's pixel lost visibility. Ad networks started hiding campaign data.

All of this happened at once. Attribution went from incomplete to useless.

Before 2021, attribution wasn't great. You'd lose 10 to 20% of your data. Phone calls, offline orders, direct traffic through obscure channels. But you could see most of it.

After 2021, you lose 40 to 60% of your data. Maybe more. Your platforms tell you one story. Your business tells you another. And you don't know which is real.

That gap isn't shrinking. It's growing.

Why Attribution Tools Can't Fix a Data Problem

Attribution vendors built billion-dollar companies on one promise: we'll solve the missing data problem.

They can't. No vendor can.

Here's why. Attribution is broken because the data is broken. Apple, Google, and Meta don't give you the data anymore. No tool can create data that doesn't exist. You can't mathematically invent what you can't measure.

So what do attribution tools do? They guess. They use statistical models. They apply algorithms. They fill the gaps with predictions.

Predictions are fancy guesses. They're not facts.

Founders spend thousands of dollars per month on these tools. They get prettier dashboards. Shinier reports. More colored boxes. But the underlying problem stays the same: incomplete data.

They're treating a data access problem like a math problem. You can't math your way out of information you don't have.

The Paradox: More Spending Equals Less Knowledge

Here's the cruel irony of modern attribution.

The more you spend on attribution tools, the more you increase your reliance on guesses. The more confident you feel in your guesses. The worse your decisions become.

You buy an attribution tool for $500 per month. It shows you attribution across 5 channels. It looks authoritative. You trust it.

But it's model-based. The model is 60% accurate at best. You're making million-dollar budget decisions on a tool that's wrong four days out of every ten.

And you don't know it's wrong until you look at your actual financial results. By then, you've already wasted three months of budget.

The founder who skips the tool and looks at their bank account isn't as fancy. But they're right more often. And they know the difference between what they know and what they're guessing.

The Alternative: Accounting-Based CAC

Here's what works instead.

Forget attribution models. Forget platform data. Use accounting.

Your bank account doesn't lie. Your credit card statements don't guess. Your QuickBooks data reflects reality.

Customer acquisition cost is a simple math problem:

Total Expenses (from your P&L) / New Customers Acquired = CAC

Not just ad spend. Not just marketing. Everything on the P&L that touches customer acquisition: ad spend, payroll, software, tools, overhead. All of it.

Your accounting system shows you the total expenses. Your CRM or ecommerce platform shows you the customers. The division is your real CAC.

This number doesn't care about iOS 14. It doesn't rely on platform estimates. It's built on the two facts that matter: total money out, customers in.

How All-In CAC Actually Works

The problem most founders face isn't calculating CAC. It's deciding which customers count.

Did the customer from three months ago count against this month's budget? What about deals that took six months to close? What about customers who came back because of a referral?

These questions tie founders in knots. So they give up and buy an attribution tool.

Here's the simple version that actually works.

Calculate an All-In CAC using a rolling 90-day window. Take all expenses from your P&L for the last 90 days. Divide by all new customers closed in that 90-day window.

That's your CAC. Not perfect. Perfect is impossible.

But it's real. It reflects actual cash out. It reflects actual customers in. It removes the noise of monthly fluctuations while staying recent enough to be useful.

Most founders we work with have an All-In CAC between $1,200 and $4,500. If yours is over $5,000 and your average deal is under $10,000, something's broken. Time to investigate.

The $2,800 vs. $181 Story

We audited a $9M revenue business. Mid-market distribution company. Good margins. Stuck growth.

Their CFO said CAC was $181. That's what the platform data showed.

We did the math using their actual accounting records. Bank statements, invoices, customer acquisition dates.

Real CAC was $2,800.

That's a 15 to 1 difference. Not 10%. Not a margin of error. A completely different number.

Why? The platform data was only capturing 30% of their customer acquisitions. Phone sales, repeat customers, channel partners, inbound leads with no UTM data. All invisible to the platforms.

The $181 CAC made them think everything was working. Spend more! Grow faster!

The real $2,800 CAC told them they were bleeding money on inefficient channels.

They cut spending in the channels that "looked good" on platforms but cost $2,800 per customer. They doubled down on the channels the platforms couldn't see. Revenue grew 40% in six months. CAC dropped to $1,400.

All because they stopped trusting platforms and started trusting their bank account.

The Attribution Replacement: Red, Yellow, Green

We built a simple framework that replaces attribution entirely.

Instead of attributing revenue to channels, you rate vendors and channels by their CAC efficiency.

Green: CAC is below your break-even threshold. Spend more here. Scale it.

Yellow: CAC is between break-even and acceptable. Monitor it. Don't cut it yet.

Red: CAC is above what you can sustain. Kill this channel or redesign it.

That's it. Three colors. One number per channel. No attribution models. No guesses.

You get this data directly from your accounting records, your CRM, and your payment processor. Not from platforms. Not from tools. From your actual business data.

Most founders can build this scorecard in an afternoon. Update it monthly. Make decisions in 15 minutes.

How to Switch from Attribution to Accounting

You don't need a massive project. You need four things.

Step 1: Define your acquisition window.

Choose 30 days, 60 days, or 90 days. We recommend 90. It's recent enough to be useful. Old enough to smooth out weekly noise. Stick with it every month.

Step 2: Export your expenses.

Pull your total expenses for that window from your P&L or accounting system. Include ads, payroll, tools, events, partnerships, overhead. Every dollar on the P&L that touches customer acquisition.

Step 3: Export your customers.

Pull the list of customers acquired in that window from your CRM or accounting system. Include customer name, acquisition date, and deal value. You need to know how many customers came in during that period.

Step 4: Divide and check.

Total expenses divided by number of new customers equals CAC. Write it down. Check it monthly. If it's going up, something changed. Investigate what.

That's your system. No tool needed. No attribution model. Just math from your actual business data.

What You Gain When You Stop Chasing Attribution

Clarity. Time. Money. In that order.

Clarity: You know what your customers actually cost. Not guesses. Not estimates. Facts.

Time: No more wrestling with attribution tools. No more waiting for reports. You have the answer in minutes using a spreadsheet.

Money: You stop spending on false positives. You stop cutting channels that actually work. You grow more, spend smarter.

Most founders we work with find they're overspending on 2 to 3 channels by an average of 30 to 40% each. That's tens of thousands of dollars per year they didn't see because the platforms were lying.

When you use accounting instead of attribution, you find that gap. Fast.

Frequently Asked Questions

What if I don't have good CRM data?

Most founders don't. Start with what you have. If your CRM is bad, your sales process is worse. Fix the CRM first. In the meantime, count customers from your accounting records. It's not perfect, but it's better than guessing.

What about indirect customers or channel partners?

Count them. They came in during your acquisition window. They count toward your marketing efficiency. If partners are dragging your CAC up, that's useful information. You can adjust the deal or the partnership.

How do I handle customers who take 6 months to close?

Choose a consistent rule and stick with it. Use the close date, not the lead date. That's when the customer actually arrived in your business. Be consistent month to month. Consistency matters more than perfection.

What if my CAC is really high?

That's information. It means your acquisition engine needs work. You're either targeting the wrong customers, spending too much to reach them, or both. The good news: knowing the problem is half the solution. Now you can investigate.

Can I use this framework if I have 100 different channels?

No. You'll go crazy trying to track all of them. Focus on your top 5 to 7 channels. If a channel doesn't add measurable volume, kill it. Simple wins.

How often should I update this calculation?

Monthly. Ideally the first week. It takes 15 minutes. You see trends faster. You catch problems sooner.