Measuring Customer Acquisition Cost at Early Stage Startups

Measuring how much it costs to acquire customers is essential for scaling a business, but it’s hard at the earliest stages. Here I want to explore how to do it.

Seed stage companies still have a lot to build. Your product will be built before you get revenue from it, so fundraising covers the difference. This is a good reason to raise VC, because the product can create a lot more value than the cost of building it.

Similarly, if you want to scale, you need to spend money to acquire customers that should generate far more revenue than you spent. You can bootstrap, only spending revenue from past customers to get new customers, but a competitor can scale faster by front-loading that spending. To justify your fundraising amount, you need to explain the acquisition model.

These two asks, for building and scaling, are similar:

  • I need $X to hire Y people to build Z product in M months.

  • I need $X to acquire Y users to reach Z revenue/scale in M months.

You know how much engineers cost. But how much do your users cost?

Measuring Customer Acquisition Cost (CAC) is a fundamental marketing metric. I wrote an introduction to CAC here. Start there if you’re new to marketing, and then continue reading back here.

A sophisticated company will have very established CAC metrics. We’re lucky that some companies have used their metrics as content marketing. Meta, I know. Rippling released their Series A fundraising memo which shows many related metrics. Go here and download it.

They include an example of CAC, and one trending in a great direction:

Rippling the company has engineering and marketing and sales teams to build out all these metrics.

What about a Seed Stage Startup?

Maybe it’s just the founders, so what should they do to invest in these metrics?

Every ideal is a judge, so comparing your seed stage metrics rigor to a Series A or B company can hurt. But it will help you get sophisticated. I routinely see seed stage pitches which don’t have these metrics, not even an attempt. Here are many obvious reasons:

  • Founders doing the work, which seems free.

  • Friends using the product, which muddles when the customer was acquired.

  • Difficult to track tactics, like customer referrals.

  • Low volume means noisy metrics.

  • Low budget means some channels (scaling ad spend) can’t work.

These are all reasonable, but don’t let that stop you from trying. Here are some ideas on what to do. I’d love to hear more if you have ideas here. Email me and I’ll add to this post:

Count Your Salary

Let’s say you’re doing sales or writing content. Surely the CEO’s salary shouldn’t be counted in your CAC? Yes, it should, but ask “what would this cost to 10X?”. It would mean hiring more people. What would you pay them? Use that as a proxy for the cost of this activity. The on target earnings of your sales reps can be calculated the same way, where you assume a commission would be paid.

Discount Your Effectiveness

Founders are the most effective sales reps. Many technical founders find this hard to believe, hoping to solve sales with hiring. That won’t work, and you need to assume the opposite. Customers love hearing from founders who have agency over the product. One reason companies tolerate early stage startups is that the service level from founders is so high. However much you’ve closed, discount that rate for new reps.

The same is true for other channels, like content marketing. People that are hired to write are not experts in your vertical, and they need to spend more time studying the space. If you’re a founder writing about what you’ve experienced, you’re more productive.

Measure Time

If you can time box your efforts, you’ll know what to modify when estimating productivity and salary. Very many marketing channels are dollar cheap but hours expensive. For example, you might attend a conference for $200, but it actually cost you days of prep and execution. Content is all about consistency over time, which is a large investment.

Diminishing Returns vs Compounding Returns

Projections are hard because rates change. If your product has social invites, for example, the conversion rate of the 5th invite someone receives is literally 40X+ that of the first. The reason is that more social proof builds trust. The same is true for a newspaper trying to convert a visitor to a subscriber. Your 5th visit demonstrates more proof than the 1st. More is better.

But other things get worse with scale. If you get more traffic from search engines, it might fall outside your ideal customer profile, and conversion rates drop. You can’t just project out that growth rate. Ad targeting is similar, where going further out reduced effectiveness.

You don’t want to successfully fundraise on rosy projections and then miss your goals. Down rounds are worse than initial lower prices. Plus good investors know everything in this post and will discount your numbers. Include realistic diminishing returns where appropriate in your projections.

Run Experiments Now

Very many companies defer work on customer acquisition because they want to focus elsewhere. Maybe you have poor retention so you don’t want to spend money acquiring customers that will churn. It’s fair to work on core product and retention first. But the problem here is time. Very often learning about what marketing channels work takes experimentation, and experiments take time to run. If you wait until you’re ready, you’ve lost that time. My recommendation is to spend at least some time thinking about and measuring this. Time box the work and the budget. That box is helpful too because it will help you calculate input and output to measure CAC.

When a founder brags “We got all these users but spent $0 on marketing”, they think they are awesome. But actually I always think “I wonder how much bigger you’d be if you spent more”. And then I often think “Your ego is getting in the way of your management“.

It’s much more reasonable to say “We tried paid acq but the free social channels proved far lower CAC”. This actually perfectly matches the early Dropbox story, where early google ad experiments cost 100X the referral program.

Spend the First Purchase

This idea comes from Casey Winters, who has an incredible blog. After customers make an initial purchase, set that as your budget for performance marketing. The reason is that the data is crisp, including on lifetime value (LTV). A lower bound of  LTV is this initial purchase, and you want your CAC to be less anyway. If your campaign can produce the same sales volume, you can scale with more confidence. I also like that it’s such a simple rule of thumb, allowing you to move fast without analysis paralysis.

What Else?

I’d love to hear more tactics here to help improve the quality of early stage metrics. Send me an email at or reply here.