I recently wrote about the most common mistakes with CAC (customer acquisition cost) that can derail growth efforts before you even get started because CAC is a metric that's foundational to growth strategy.

In this post, we'll look at the faulty myth of "Average CAC," and the most meaningful CAC segmentations that you should be paying attention to instead.

DON’T RELY ON AVERAGE CAC

CAC = Total Marketing + Sales Expenses / # of New Customers Acquired

The above basic CAC calculation will get you the average CAC across your business. For internal operations and decision-making however, average CAC is almost always useless.

Averages hide extremely important insights that could either help or hurt you. Making CAC truly useful as an operating metric comes down to segmenting. There are four major ways you can and should segment, as well as a few unique cases we’ll cover at the end of this section.

1. Segment Based On Customer Type/Attributes

Most companies have different types of customers. Different types of customers behave very differently. They may be acquired by different types of channels, cost different amounts within the same channels, have different conversion rates, and much much more.

The DocuSign Example – Segmenting By Customer Type

(Note: All data is 100% fictional. I'm just using DocuSign as a tangible example)

Let’s take DocuSign as an example. They have four different pricing tiers targeted at different types of customers:

As shown on their pricing page, their tiers cover the following:

Personal $10/month – Likely targeted at individual professionals, sole proprietors, contractors and very small businesses.
Standard $25/month – Likely targeted at Mid Market companies.
Business Pro $40/month – Likely targeted at companies with power users (i.e. Law Firms).
Enterprise – Likely targeted for very large companies like the Fortune 1000.

In the below spreadsheet we have some example data comparing Average CAC to a Segmented CAC.

Note: To adapt any spreadsheet for your own use, click here to open the original, then File > Make a Copy.

The calculations above assume a few things:

Marketing expenses are spread evenly across customer types
Sales expenses are attributed to Enterprise Customers. We assume the other tiers all are self service and don’t require sales.
Average marketing/sales cycle is less than 30 days.
20% of total support costs are dedicated towards supporting the free trial.

As you can see there is HUGE difference between our Average CAC calculation and our segmented CAC calculation. The Average CAC is useless when it comes to making operating decisions.

In this example we may compare our Segmented CAC to our Segmented LTV and find that the Personal Tier was not performing well. We could then evaluate a few different paths:

Do we make pricing changes to that tier?
Are there certain channels dedicated to attracting customers of that tier? If so should we cut them?
Do we need to do better education of the Personal Tier to move them up to a higher tier?

DocuSign is just one example, but every business typically has different types of customer types. Here are some more generic examples by company type of how you may segment your CAC on customer type:

SaaS – A lot of SaaS products serve multiple segments of the market: Very Small Businesses, Mid Market, Enterprise. You’ll also typically have different tiers of the product that serve these different markets. For example, HubSpot has a Basic, Professional, and Enterprise tier. If you have something similar, you want to segment the CAC of a Basic, Pro, Enterprise customer.

Freemium – In a freemium product like Dropbox, you have customers who are individuals and customers who are teams. The CAC to acquire an individual and the CAC for a team are likely very different.

Ecommerce – Vistaprint is a massive ecommerce business serving SMBs. But the SMBs range from restaurants, to small shops, to service agencies, and many other business types covering a variety of industries. Each industry behaves very differently on the CAC and LTV side and therefore should be segmented to inform operating decisions.

Here is a short list of customer attributes you may want to segment by:

Size of Company
Industry
Tier of Product
Power Users vs Casual User

Segmenting based on customer type and comparing to a segmented LTV could lead you to a number of important decisions around your business:

To focus on one customer type over another.
To change your channel strategy for certain types of customers.
To change your pricing tiers for certain types of customers.

To segment your CAC based on customer type you need to

Define each customer type.
Allocate marketing expenses to different customer types. For example, you may have certain marketing channels dedicated towards acquiring certain customer type. If a channel covers multiple customer types, then you need to split the allocation. Same goes for allocating people/salary expenses from marketing.
Allocation sales expenses to different customer types. A lot of companies have sales teams focused on different customer segments (SMB vs Enterprise or Industry A vs Industry B).

Learn Growth, No Fluff. One Essay Per Post.

2. Segment Based on Channel

The second most important way to segment CAC is based on channel. Before you say “I know, I know” and skip onto the next section, there are some nuances worth reviewing.

Different acquisition channels will produce very different CAC numbers since the costs, conversion rates, and other factors will vary widely between channels.

Note: To adapt any spreadsheet for your own use, click here to open the original, then File > Make a Copy.

The issue with this segmentation is that it depends on last click attribution. We aren’t going to cover attribution models in depth here, but we should look at how it affects your CAC analysis.

Last click attribution is fine if you have a very short (almost instant) decision funnel for a product. In these cases, the last click is likely doing a high percentage of the convincing. But if you have a longer funnel, it can lead you in the wrong direction as it takes multiple marketing touch points from various channels to convert to a customer. Something as simple as comparing CAC from first touch vs last touch attribution will produce very different results. If you do not have a blended attribution model in place, our best suggestion is to look at CAC from “multiple angles” to make channel decisions.

3. Segment Based On Geographies and Location

The location of the customer can have a large effect on your CAC. At a minimum you should be looking at CAC segmented based on country. But for a lot of companies, especially location based mobile apps, marketplaces, etc looking at CAC on a location by location basis is key to making operating decisions and gathering new insights.

4. Other Segmentation Methods

The most common/impactful ways to segment CAC are above. But here are a ton of other scenarios where we are trying to make an operating decision and need a different segmentation.

Two examples of scenarios requiring additional segmentation:

Bidding Method – In paid acquisition there are typically a lot of different ways to bid. Facebook for example has oCPM, oCPC, oCPA, etc. Segmenting CAC based on the bidding method used can tell you what bidding method produces the lowest cost customers.

Content Category – With content marketing at scale, segmenting your CAC based on the category of content where that lead was originally acquired can indicate what additional content you should produce.

VIEWING CAC OVER THE LONG TERM

One of the other key mistakes we see is looking at CAC in the short term and not understanding the long term context surrounding the number leading to bad decisions.

The key with any quantitative number is to add qualitative information and a dose of your own intuition to come to a final conclusion. Here are a few common scenarios where the numbers don’t provide all of the information.

Channel-Based Evolutions

Different acquisition channels evolve differently and as a result so does CAC based on that channel. CAC may look large in the short term leading you to cut that channel off, but if you had taken a long term view you would end up with a different conclusion. Here are two different examples.

CAC For Paid Acquisition

One of the pros of paid acquisition is that it requires less people (and therefore salary costs) to scale than other channels like content marketing. That means your fully loaded CAC number in the experimentation phase of paid acquisition might look unreasonable, but that if you scaled the channel out your CAC would be fine.

To illustrate this, we’ll look at the below example. We’ve assumed the following things:

100% of our marketing efforts is focused on Facebook Ads.
We have a VP Marketing making $170K, and a Paid Acquisition Manager making $80,000.
No Sales team (consumer company)
We are testing Facebook Ads as a channel.
Our target CAC is $25

Note: To adapt any spreadsheet for your own use, click here to open the original, then File > Make a Copy.

You can see in Jan through March we are testing Facebook Ads with a small budget. CAC ends up being WAY ABOVE our target of $25. This is because at a small scale salaries contribute a large % of CAC (more about "fully loaded CAC" here). We might look at that data and make the incorrect decision to shut off Facebook Ads.

However, if we project out and assume that we can scale the budget to $25,000 per month our CAC falls below the target of $25 even assuming that our spend gets less efficient. A great paid acquisition manager can typically manage $500K – $1M in monthly budget on a channel.

This is one of those few situations where looking at Non-Fully Loaded CAC might be useful. Either way you want to go through the exercise of projecting out to get a more accurate picture.

CAC For Content Marketing

CAC for content marketing evolves differently than paid acquisition. Two components that make content marketing different:

Content marketing is much more people-driven than paid acquisition and therefore salaries play a bigger role in CAC. A single content marketer has a ceiling on the number of quality pieces they can produce within a time period. So if your strategy is to increase quantity of content, that typically requires increased headcount.
The most successful content strategies are SEO-based and therefore get cheaper over time as a piece of content ranks higher, brings in sustained traffic (without extra effort) and therefore increased leads and customers.

Let’s look at the second point in a little more detail. In the below example we assume a few things to simplify the point:

100% of our marketing efforts is content marketing.
Average sales cycle is less than 30 days.
We have a VP Marketing at $170K, 2 Content Creators making $80,000.
Each content creator produces 5 pieces of content per month.
In the month a piece of content is new it gets 2000 visits. Each month thereafter gets 1000 visits due to SEO.
2% of our traffic converts to leads. We close ⅓ of our leads to new customers.

In the example you can see how CAC starts high in the first few months, but then steadily declines over time. The reason is that pieces of content from previous months still attract traffic without additional investment in marketing, resulting in more leads and lower CAC over time.

In the above example, if the company had a target CAC of $150, and was trying to decide if Content Marketing was worth pursuing after month 2, they would likely come to the conclusion of no. This is why you must look at CAC on how it evolves over time.

More specifically, when starting content strategies you have to:

Give it at least 6 months to see real results/metrics.
Make SEO a key part of your content strategy.

Network-Effect Based Evolutions

Products that are based on marketplaces or network effects also evolve in a unique way. Typically CAC (for supply or demand) can start high, but then decrease over time as there is a higher volume on both sides of the market and therefore liquidity.

For example, a new customer for Uber would likely be defined as a new rider who completes their first paid trip. If Uber launches in a new city, they might acquire a new download but since the city is new and there aren’t that many drivers that new potential rider can’t find a driver. This is going to drive overall CAC up in the early days of that city.

This is why most marketplace companies end up needing to raise capital to fund high CAC as they gain liquidity in the marketplace.

Saturation-Based Evolutions

The previous examples are all situations where CAC looks high in the short term, but when you play it out it decreases in the long term. Saturation on the other hand makes CAC look lower in the short term compared to how it will evolve in the long term.

Saturation happens both on a macro level where competition increases in a channel, which increases costs as well on a micro level where you get closer to the ceiling of your target audience within a channel. In both cases, costs will increase and you need to anticipate this when thinking about CAC over the long term during goal setting, modeling, and fundraising.

Conclusion

To wrap this up, I hope you can see from the examples in this post that average CAC is a misleading concept that requires a little bit of nuance in order to be actually useful to your growth strategy.

Additionally, CAC is fluid over time, and you should therefore expect channel-based evolutions, network effect-based evolutions, and saturation-based evolutions to (sometimes dramatically) impact this very important metric.

Don't be fooled by average CAC, or by the assumption that costs are static, or you might end up with some big surprises and blind spots when planning and executing growth.