Understanding the cost of revenue

While revenue maximization is a critical part of every product development, you shouldn’t neglect the second part of the equation — the costs.

The costs you acquire to generate revenue impact:

• Profitability — High revenue is irrelevant if it’s all consumed by various costs
• Payback period — Sometimes it’s even better to get less revenue with quicker turnaround
• Roadmap prioritization — Building an optimal revenue and cost structure should be one of the inputs in deciding what to do next

In this article, you will learn what the cost of revenue is, the different types of costs, and how you can use this information to make better product decisions.

What is the cost of revenue?

The cost of revenue refers to the costs accrued to generate revenue from a specific use case. It’s usually split into two groups of cost:

• Cost-to-serve — How expensive is it to serve the value for customers?
• Cost of acquisition — How expensive is it to gain new customers?

Understanding these costs will help direct your product strategy moving forward.

What are the different types of costs?

To better understand the cost of revenue, let’s unpack the two main drivers of cost:

• Cost-to-serve
• Cost of acquisition

Cost-to-serve

Cost-to-serve includes all the costs you acquire to properly serve one paying customer. These costs scale with the number of customers you serve.

Let’s take Figma as an example. To serve one customer, they need to cover the costs of:

• Customer support (the more customers = the more request to CS = the more CS people you need to pay for)
• Server space (the more customers = the more projects = the more space you need to host these projects)
• Analytics software (Figma uses various analytics tools to get insights about their users, and the more users, the higher the cost)

The most common types of cost-to-serve include:

• Physical products — The cost of building and shipping something physical to customers (e.g., Dollar Shave Club delivering their razors)
• Logistics costs — Some products acquire logistics costs, such as Uber Ears, which incurs the cost of delivering the food to customers. Although it ultimately offsets the costs to users, it still impacts its profitability
• Storage and hosting — The cost of storing user data (like Figma storing assets in the cloud) or hosting user data (like Webflow hosting websites for customers)
• Customer support — Although often neglected, as your customer base grows, your customer support team usually needs to grow equally fast, and that’s not cheap
• Tools and programs — Any third-party tools you use to serve your customer directly (e.g.chatbots) or your business needs
• Miscellaneous — Any other cost you acquire with each new paying customer. For example, the cost of sending personalized Christmas gifts every year

Cost-to-serve allows you to calculate your margins, that is, the total earnings after offsetting the cost:

Margin = revenue – cost-to-serve

If you charge your customers $20 a month for a subscription, and your average cost to serve one customer for a month is $13, then:

Margin =$20 – $13 = $7

This means that you generate a $7 profit with every renewal.

You can also calculate the margin percentage:

Margin percentage = margin/revenue

In our example, it would be:

Margin percentage = 7 / 20 = 35 percent

Here you capture 35 percent of your revenue for further investments.

Cost of acquisition

Cost of acquisition, or customer acquisition cost (CAC) includes costs related to getting a new customer.

Although it’s often simplified to marketing and advertising costs, there’s much more to that.

Let’s take Netflix as an example. The journey to become a paying customer looks like this:

Users first come to the platform from referrals, organic search, or Netflix’s paid ads. Part of Netflix’s growth model involves first serving a free trial for newcomers to get them hooked.

This free trial is also an additional cost that Netflix needs to acquire to obtain paying customers.

So, if the average cost of getting users from ads is $0.50, and the cost to serve the trial for a week is roughly $2. Let’s assume the following scenario:

• Four users come to the platform from the advertisement
• Two of them start a free trial
• One of them becomes a paying customer

The actual cost of acquiring this particular customer would be $6:

Acquisition cost = 4 users  x  $0.5 ad costs + 2 trials  x  $2 = $2 + $4 = $6

But it’s still only a part of the equation, you can also have these other types of acquisition costs:

• Performance marketing — Any sort of paid advertisement is a cost you acquire. You get new paying customers on the product
• Cost-to-serve before conversion — Just like Netflix’s free trial example, cost-to-serve includes all costs required to serve users before they convert
• Referrals — If you incentivize users to invite more people to the product, for example, by giving a 10 dollar voucher for each successful referral, that 10 dollars is part of the cost of acquiring that new customer
• Partnerships — Any paid partnership, such as a complementary product recommending their users to try out your product, is an additional acquisition cost
• People costs — Your sales and marketing teams also generate costs that are often fully focused on acquiring new users for your product
• Tools — If you use expensive acquisition-focused tools, that’s also a cost worth including. For example, Salesforce can become quite an expensive investment if you have a big sales team
• Miscellaneous — There might be other costs worth including. For example, paying social media influencers to talk about your product or paying for guest posts on renowned blogs

Ultimately, while marketing teams often use CAC to calculate their performance marketing costs, when you look at this from the product perspective, it’s often a significantly higher number:

Cost of acquisition = all costs required to get one paying customer

Attributing all these costs to a particular customer is close to impossible, so we often resort to using averages.

For example, if in May you acquired a total of $30,000 in acquisition-related costs, and you got 5,000 new customers, you can assume:

Cost of acquiring one customer = $30,000 / 5,000 = $6

Using cost-to-serve and customer acquisition cost to calculate a payback period

After understanding your customer acquisition costs and cost-to-serve, you can derive your payback period:

Payback period = acquisition cost / (revenue per period – cost-to-serve a period)

For example, say you sell a monthly subscription plan for $30. It costs you roughly $15 to serve the value proposition for each month.

Let’s assume that your conversion funnel looks as follows:

1. You get ten users on the platform through ads for $5
2. Five of them start a two-week free trial (5 users x $15 cost-to-serve x 50 percent for only half the usual period).
3. One of these trials allows the user to convert to a paying customer.
4. Your other acquisition costs (tooling, people, etc.) average roughly $10 per new customer

That gives you a $15 cost-to-serve and a $52.5 acquisition cost ($5 for ads + $37.5 to serve trials + $10 for other costs)

Payback period = 52.5 / (30 – 15) = 52.5 / 15 = 3.5 

After 3.5 months, you break even, which means you offset all costs related to acquiring the customer and generate your $15 profit margin every month (revenue – cost-to-serve).

Why does this matter, though?

The payback period might be even more of a factor than your profitability and revenue per se.

Let’s visualize that with an example. Say you serve two different use cases to your customers. Each type of customer sticks with you for roughly one year before churning, and you bill them a monthly fee.

It costs you $500 to acquire a customer for use case A and $30 per month serving them. The revenue you get from these customers is $100 a month.

For use case B, you acquire a $200 acquisition cost and $10 monthly cost-to-serve. You charge them $50 a month.

When we look at these use cases from the financial perspective:

• Use case A:
  • 500 / (100 – 30) = 7 months payback period
  • 12 x 100 – 12 x 30 – 500 = $340 total profit in a year
• Use case B:
  • 200 / (50 – 10) = 5 months payback period
  • 12 x 50 – 12 x 10 – 200 = $280 total profit in a year

Although from a pure revenue perspective, the use case A brings you more money, there’s one caveat.

The revenue-cost structure of use case B allows you to recover your costs two months faster. As a result, you can reinvest in new acquisitions sooner. That’ll mean more new customers, which will also pay back sooner, allowing for yet another round of reinvestment.

How to use the cost of revenue to make better decisions

Now that we’ve covered what cost-to-serve, acquisition cost and the payback period is, let’s take a look at a few examples of how you can use these data points to make better product decisions:

• Use case assessment
• Better understanding of tradeoffs
• Evaluating acquisition source

Use case assessment

Just as we covered in our payback period example, analyzing how expensive it is to acquire users and serve specific use cases gives valuable input on profitability and potential growth rate.

Use this input and plan your next steps to reach a desired balance between:

• One-year revenue
• Lifetime value
• Reinvestment rate
• Segmentation

Better understanding of tradeoffs

Many decisions you make have consequences on your revenue and cost metrics. Let’s take the most common example — monthly versus annual subscription plans.

Yearly plans usually offer an instant payback period (as you collect all revenue upfront) at the price of reduced total revenue potential (yearly plans are usually discounted). In contrast, monthly plans tend to generate higher total LTV at the cost of longer payback.

Combining that insight with the difference they make on new vs. recurring revenue will help you make well-informed decisions and predict their impact on the long-term revenue structure.

Ultimately, the more nuisance factors you consider when making the decision, the smarter the final decision tends to be.

Evaluating acquisition source

By segmenting your acquisition costs, cost-to-serve, and payback periods per different acquisition sources, you might be able to identify interesting outliers.

It might turn out that users from your most expensive acquisition source tend to buy long-term subscriptions, thus still offering a quick payback. As counterintuitive as it might sound, that would be a powerful signal to invest heavily in that most expensive channel if you prioritize quick reinvestments to fuel the growth loop.

Similarly, the fact that a specific acquisition channel is cheap doesn’t necessarily mean it’s a good channel. If it takes a year for users to come from that channel to eventually pay back (say, because they chose the least profitable use cases), investing heavily in this cheapest channel might lead to freezing your resources for a long period, which might even kill the whole business.

Segmentation

You can also analyze the cost structures per various user segments, such as segmenting by:

• Frequency of use
• The country you acquired them from
• Feature usage

Proper segmentation often reveals surprising insights. There might be a segment of heavy users with incredibly high cost-to-serve. That signals you might rethink your pricing approach or offer a separate, more expensive plan for heavy users.

You might discover a segment that uses a specific feature heavily, but that feature generates high cost-to-serve, which eats all your margins. Maybe you should price additionally for that particular feature to offset the extra cost? Or, if used rarely, maybe you should kill the feature to reduce your overall costs.

If users with a specific profile or from a particular region have significantly lower or higher acquisition costs, you might rethink your acquisition strategy and product direction to focus on attracting low CAC users.

Closing thoughts

Before you use customer acquisition costs, cost-to-serve, or payback period as input to your decision-making process, ensure you understand how these are defined and how they impact the subsequent analysis. Sometimes you might need to tweak the formula here and there to get the answers you care about.

Whenever someone talks about cost-related metrics, my first question is always, “What’s included and what’s NOT included in our CAC/cost-to-serve definition?” That helps in avoiding painful misunderstandings.

Featured image source: IconScout