This is not a story about one company. It is a pattern I have watched repeat across twenty years of advisory work, from seed-stage startups burning their first round to VP roles inside SaaS companies at the eight-figure ARR mark.
The pattern looks like this. A growth team shows me their CAC dashboard with the kind of pride people reserve for newborns. The chart is beautiful. Down and to the right. Twelve months of steady improvement, blended CAC payback under ten months, ratios that would clear a board meeting.
The business is failing. Pipeline is up. Revenue is flat. Sales cycles are getting longer. And the CFO has quietly stopped trusting the marketing reports.
The dashboard is not lying. It is answering the wrong question.
Every time I get a call from a startup asking if I can help them map out their marketing stack before they build it, I feel genuine relief. Because I know that at least one more company is not going to experience that slow death over a three-year period. And it is always three years. That is how long it takes for the scrambled egg to become undeniable.
1. How the scrambled egg gets made
The pattern almost always starts right after a first funding round.
The company is convinced it is the next unicorn. The founders ask executive friends what they should do. Those executive friends, many of whom have quiet referral arrangements with software vendors, recommend the tools that earn them their ten percent. Before the company knows what happened, they have a HubSpot or a Salesforce they are not ready for, connected to twenty other applications, three reporting platforms, and a second CRM someone added specifically for tracking customer feedback.
Nobody set out to build a broken system. The marketing team hired at a price point that reflected the budget rather than the complexity of the problem. The data team did the same. Both groups set up their systems in isolation and then tried to connect them with middleware. The PPC ads start showing excellent conversions and a cost per lead that looks healthy in the dashboard. On the sales side, nothing is closing. Without a clear attribution path connecting those two realities, a company can spend millions on advertising and generate nothing for it, and the dashboard will not tell them for months.
I have seen this at companies with two people in a room and companies with two hundred. The stage that is most dangerous is right after first funding, when the team has enough money to buy enterprise tools but not enough experience to implement them properly, and enough confidence to resist outside help until the problem is too expensive to ignore.
2. The diagnostic question I ask in the first fifteen minutes
When I get on a call with a company that I suspect is in this pattern, I ask one question directly: how do you connect your data across your marketing and sales stack?
The moment I hear Zapier or Segment, I have concerns. Not because those are bad tools. They are genuinely excellent tools. Segment in particular is one of the most powerful customer data platforms available. But both tools give a relatively inexperienced team the ability to build a data architecture of tremendous complexity with very little friction, and complexity without expertise is how you get a scrambled egg.
In the wise words of Uncle Ben: with great power comes great responsibility. A junior data analyst with access to Segment and twelve SaaS integrations can build something that looks impressive and measures nothing useful, and it will take eighteen months before anyone realises the reports have been wrong the entire time.
The answer I am hoping to hear is something simple. We use one CRM, our ad platforms push data into it via native integration, and we run reports directly from that one source. Simple systems, properly implemented by people who know what they are doing, outperform complex systems every single time.
3. What CAC actually measures and what it does not
Most growth teams calculate CAC as total marketing spend divided by new logos closed in a defined period. That is customer acquisition cost in the same sense that miles per gallon is the cost of driving. It is a useful input. It is not the answer.
Four costs are typically missing from the calculation:
Sales loading. Most CAC formulas under-count sales. They include AE base salaries if at all, but rarely include sales engineering, fractional sales leadership, the SDR cost of qualifying out, or the loaded cost of demos that do not convert. In B2B SaaS, the actual sales contribution to CAC is often one and a half to two times what the dashboard shows.
Post-sale customer success cost. A customer you closed but could not onboard, who churns at month six, has a higher real CAC than a customer who stayed for twenty-four months. Most dashboards do not reflect this until far too late.
Retention failure tax. When churn rates are higher than the model assumed, the implicit CAC of every retained customer goes up. Growth teams usually wait for the year-end LTV review to absorb this. By then it is six quarters into the math being wrong.
Segment dilution. The biggest one. CAC blended across segments hides the segment where the math is actually working. The smaller, better segment subsidises the larger, worse one in the average. The team optimises the average and slowly kills the segment that was holding everything together.
You can argue that some of these belong in LTV rather than CAC. The point is not where the line goes. The point is that when a CEO asks what does it cost us to get a customer, the number on the dashboard is consistently thirty to sixty percent low.
4. The case for segmented, fully-loaded CAC
I worked with a B2B SaaS company at the fifteen million ARR mark during my time advising through VonClaro. CAC payback on the dashboard was fourteen months. Excellent. Healthy. Board-ready.
When we segmented:
- Segment A, the small one: CAC payback approximately nine months, net revenue retention approximately one hundred and twenty percent, gross margin approximately seventy-eight percent
- Segment B, the large one: CAC payback approximately twenty-two months, net revenue retention approximately ninety-five percent, gross margin approximately sixty-two percent
Marketing spend was distributed roughly seventy percent to Segment B. That allocation, combined with segment-blind averaging, was masking the fact that one segment was profitable and the other was a slow leak.
Once we re-tagged spend by segment, loaded in the full sales and customer success costs, and rebuilt the dashboard around segmented CAC payback paired with retention, two things became obvious that had not been obvious for eighteen months. Most of the marketing budget was buying customers who were never going to be profitable. The segment that was working was systematically underfunded because nobody could see it.
We did not increase the budget. We shifted approximately forty percent of spend from Segment B to Segment A over one quarter. Twelve months later, blended CAC was higher on the report because Segment A is more competitive to acquire. But contribution margin per retained customer at month eighteen was up materially. Revenue rate of change had inflected. The board started trusting the marketing reports again.
The lesson is not segment your CAC. Everyone knows they should segment CAC. The lesson is that segment-blind averages will always be more attractive than segmented reality, and teams will choose the easier number until something forces them not to.
5. The metric I use instead
Here is the framework I use in my advisory practice now when working with companies past five million ARR:
Contribution margin per retained customer at month eighteen.
It is a longer name than CAC. It is a better number.
Why month eighteen. Most B2B SaaS churn pattern shows up clearly by month eighteen, especially when onboarding was lazy. Before that you are looking at honeymoon performance. After that you are looking at survivors, and you cannot run an acquisition decision based on survivors alone.
Why contribution margin. It forces you to subtract the cost of serving the customer, not just acquiring them. Sales, customer success, support, and infrastructure attributable to the segment all get loaded in. You cannot hide a poorly-margined segment inside a blended margin line.
Why per retained. Because the customers you lost in month seven are still part of the acquisition math. Otherwise you are answering what did we pay for the ones who stayed, when the actual question is what did we pay overall.
This is what decision-grade economics looks like for acquisition. CAC is directional. Contribution margin per retained customer at month eighteen is the number you make actual decisions from.
6. What changes when you fix the math
In my advisory work, when this metric gets calculated for the first time, three things happen within sixty days:
Budget allocation arguments stop being arguments and start being math. Marketing and finance have a number they can both reason about. The number does not favour either side.
At least one segment gets quietly de-prioritised. Usually the segment everyone knew was the strategic one and nobody had been willing to question.
The marketing team gets calmer. Performance metrics suddenly mean what they say, and the team stops trying to defend numbers that do not survive contact with finance.
What did not go to plan, in one case. The team I describe above lost two senior marketers in the quarter following the spend shift. The segments they had been responsible for were no longer the focus. They saw it before management did and left. That was a real cost. I probably should have managed the transition with more warning.
What I do differently now. I make the segmentation finding visible to senior marketers before the budget shift. They get to be co-authors of the reset, not casualties of it. And I ask the data and attribution question in the first fifteen minutes of every engagement, because the number on the dashboard is only as good as the system underneath it.
What this means for growth leaders
The CAC number on your dashboard is doing two jobs. It is a real-time operational metric and it is a strategic input. It is pretty good at the first job. It is terrible at the second.
If you are making decisions about which segments to fund, which channels to scale, which campaigns to kill, blended CAC will routinely point you the wrong direction. You need decision-grade economics, not directional ones. That means segmented, fully-loaded, and paired with retention.
But before you rebuild the metric, check the foundation. Ask your team how they connect their data. If the answer involves more than three systems and a middleware layer holding them together, you probably do not have a CAC problem. You have a scrambled egg. And the scrambled egg has to be untangled before the number on the dashboard means anything at all.
Cheap staff and enterprise tools do not scale. Good people and simple systems do. Pay for talent up front and save your money for the rest of your time.
Further reading
- Bessemer Venture Partners, State of the Cloud: CAC payback benchmarks by stage
- David Skok, SaaS Metrics 2.0 at ForEntrepreneurs: the foundational segmentation framework
- OpenView, SaaS Benchmarks: segment-level retention and gross margin reference
- Google Analytics 4 attribution documentation: understanding cross-channel data integrity
- Kyle Poyar, Growth Unhinged: current SaaS pricing and CAC dynamics
Run the numbers yourself. Use the free CAC Calculator at robtcase.com/tools/cac-calculator to calculate your fully-loaded customer acquisition cost, LTV to CAC ratio, and payback period. The tool includes the segment comparison logic described in this article.