Startup metrics & ratios

Customer acquisition costs (CAC) is arguable among the most important and most dangerous metrics in the startup world. It measures the marketing and sales expenses necessary to acquire one new user. CAC is one part of measuring the sustainability of your business. Without knowing it you’re spending your resources inefficiently at best and sprinting towards bankruptcy at worst.

So, it’s fairly obvious why it’s important. But why is CAC dangerous? There are multiple different definitions and flavors which all have their pros and cons. This makes it very important to know exactly what you’re measure (this is technically true for all data…) and hard to compare your CAC to that of others.

Let’s have a quick look at four different variations that exist:

Fully-loaded CAC

This includes all costs related to acquiring new customers.

Salaries for all marketing and sales-related efforts. Overhead costs like the share of your rental costs that can be attributed to marketing and sales. Costs to support users during a free trial (free trial is an acquisition channel). Costs to onboard users. Costs for marketing agencies. Costs for marketing and ad creatives. Advertising costs. Share of your IT tooling costs that are used to acquire new customers.

You get the idea.

Blended CAC

Fully-loaded as well as Blended CAC implicitly assume that all marketing effort support each other. In practice, this assumption seems reasonable. Someone who has already read one of your blog posts might be more receptive to your paid ads.

Conversely, someone who discovered your business through an ad will probably get closer to converting to a customer by attending your webinars, listening to your podcast, or reading case studies on your blog.

While the Blended CAC is relatively easy to compute — you only need all sum all your marketing and sales expenses - it does not tell you what is driving the growth. Segmented CACs are needed for that.

Paid CAC

Same as the Blended CAC, but you’re dividing by the number of customers acquired through paid channels instead of all acquired customers.

For example, customers that came in through ads on LinkedIn, Google, Instagram, or a sponsoring would be considered. Customers that came in through webinars, case studies on your blog, or word of mouth would be ignored.

The Paid CAC gives you an idea of how well your paid marketing performs.

Segmented CAC

Depending on your product model and acquisition channels it can make sense to calculate multiple CAC values across different dimensions.

For example, if you have very different pricing plans it makes sense to calculate separate CAC values for each of them. Imagine your cheapest plan (Basic) is 12.99 USD/user/month and often chosen by single users, while the most expensive one (Platinum) is 64.99 USD/user/month. The Platinum plan is often used by teams of >10 people, so the average order value is more than 50 times higher than for the Basic plan.

Certainly, your CAC for the Platinum plan will be orders of magnitude higher than for the Basic plan, rendering an average CAC useless.

You can also segment your CAC by advertising channel (Twitter vs. LinkedIn vs. Reddit), marketing campaign, geographic location of your customers, target industry your customers are from, or any other attribute that shows high variance across your customer base or otherwise skews the distribution.

Regardless of which specific variant you're looking at, CAC is an indicator of how effective your marketing and sales efforts are.

Customer success, retention and CAC

Another tricky topic is how to include costs for retaining existing users in your CAC analysis. Keeping your existing users happy and engaged plays a key role in your overall acquisition strategy. Following that train of thought, customer support and customer success expenses should also be taken into account.

In addition, you often want to not only look at your customer acquisition costs but also your CAC/LTV. Efforts to retain customers and make them stick around longer will increase your customer lifetime value (LTV). Not including these costs in your CAC will artificially increase the CAC/LTV ratio, as the costs are not captured in the CAC but the benefits improve the LTV.

A lot of SaaS companies approach it the following way: Product development costs (which can help in acquiring new customers and/or retain existing ones) are treated as R&D expenses and are not included in the customer acquisition costs.

Then they have a CAC for new users and a CAC for expansion revenue. This is revenue from up- or cross-selling existing customers. The latter one then includes costs of your customer success team or paid ads to activate and retain customers.

Alternatively, you can include customer success and other retention costs in the numerator. To be consistent you also have to add returning customers to the denominator, instead of only using those who are new.

Formula

If you’re just looking for one summary metric, then Blended CAC is the way to go — especially for most bootstrapped founders and smaller teams. Usually the Fully-loaded CAC expenses like salaries or rent costs are neglectable in those settings and omitting them makes the calculation easier.

Also those smaller companies often don’t run a lot of paid ads, making the Paid CAC somewhat useless. If you have meaningful segmentation criteria already then go ahead and split your Blended CAC up.

There are quite some caveats when calculating any form of Customer Acquisition Costs though, which is why I highly recommend going through the articles linked at the end of the CAC section.

Usually, Customer Acquisition Costs are calculated for a specific period — e.g. on a monthly or quarterly basis. When picking a timeframe, you have to be mindful of the specific characteristics of your business model.

For example, if you have long sales cycles of >90 days it doesn’t make much sense to look at a 30 day CAC. The costs that you incur now can’t be attributed to the new customers of the same month. The benefits of today’s expenses will only materialize next quarter.

Again, make sure to check out the links below.

• Cost of marketing and sales: All expenses for your marketing and sales efforts, except for salaries, overhead (rent, legal costs, snacks, recruitment, office perks etc.), and software. Here’s a non-complete list of some of the most common expenses:
  Costs for events. Paid marketing expenses. Costs to support users during a free trial (remember: free trial is an acquisition channel). Costs to onboard users. Costs for marketing agencies and other consultants. Costs for marketing and ad creatives. Costs for discounts that you give to new customers. SEO-related expenses like paying for backlinks. Influencer costs. Affiliate fees that you pay out.
• Number of newly acquired customers: Simply the total number of users that you have acquired in the given period of time.
• While we’re generally aiming for low CACs, there are a lot of cases where high costs to acquire new customers are perfectly fine — namely when the lifetime value of these customers (LTV) is even higher.
  This can be the case for expensive physical products like furniture or cars, travel booking, or a lot of enterprise B2B SaaS businesses. If you’re charging 700 USD per year per customer and they stick for 3 years on average, it’s not a big problem to spend 500 USD acquiring them. In these cases, you just have to manage your liquidity (see below).

Why it matters

• Without knowing your costs to acquire new customers you are flying blind. You have to know your CAC to estimate when you can become profitable and to know the effectiveness of your marketing and sales efforts. If you’re spending resources on getting new customers, you want to know how well those resources are spent. You also want to know whether they’re deployed more efficiently over time.
  Knowing your acquisition costs further allows you to make calculated bets. For example, you could decide to let your CAC increase by 30% to growth faster. Then you know how your breakeven point moves as well (assuming everything else stays constant).
• To avoid running into liquidity issues. You might sign loads of new customers but might have run out of money next month because your acquisition costs are too high compared to your payback period.
  While the ratio of CAC/LTV hints at the sustainability of your business, looking at the absolute CAC value can prevent you from scaling too quickly. Combined with the payback period and your liquidity, i.e. cash on the bank it determines how many customers you can acquire in a time period.

How to improve it

• If you’re running paid ads, improve your creatives and the ads’ targeting. This will help to improve the ads’ click through rate, i.e. make them more effective.
• Improve your landing pages to increase conversions.
• Since CACs are the costs to acquire customers, you can try to improve the conversion of free users to paid users.
• Create a brand and product that your customers love. They will talk about it, share it on social media, and with their friends, colleagues, and industry peers. Word-of-mouth produces more inbound leads which are significantly cheaper than outbound leads.
• Hardest to execute but with the biggest possible impact: Improve your Product-Market-Fit. The more potential customers need your product or service and easily see the value you provide, the less resources you will have to invest to turn them into actual customers.

What to watch out for

• Do not confuse Customer Acquisition Costs (CAC) with Cost Per Acquisition (CPA). As noted above, CAC specifically refers to the cost to acquire a customer. CPA can be the cost of any acquisition — a lead, trial user, sales qualified lead (SQL), user of an unlimited free plan etc.
• CAC is cost to acquire new customers, not new users. Assume you have a free plan with a 20% conversion rate from free to paid. It costs you 10 USD to acquire a user (→ Cost Per Acquisition (CPA)), then you CPA is 10 USD while your CAC is 10 USD / 20% = 50 USD. This is overly simplified since you’ll have many extra expenses that also flow into your CAC calculation, but it highlights the difference of CPA vs. CAC.
• Costs for marketing and sales don't only impact the cost to acquire new customers. They also impact the retention of existing customers. The better your onboarding, customer success, and Product-User-Fit for new customers is, the longer they will stick around.
  Meaning the better targeted your marketing and sales efforts are and the closer they already bring the user to their setup and aha moment, the longer these users will use the product on average.
• Like all startup metrics, Customer Acquisition Costs don’t remain constant. They will evolve and tend to go up over time as you have to get further and further away from your ideal customer profile. How they develop depends on the channel:
  • For organic search, content marketing, and word of mouth: The CAC will decrease over time. Your brand will become better known over time, word of mouth (hopefully) gets stronger the more people use your product, and slowly your website content will climb the search results pages.
  • For paid channels you will most likely see a decrease in acquisition costs in the beginning. That’s because as you optimize your creatives and targeting the ads resonate better with viewers and become more effective. At some point the CAC will slowly start increasing as you have to tap into audiences that are further and further away from your ideal customer and the product-user-fit decreases.
• Are you giving our referral fees or discounts to new customers? Then you have to include those in your Customer Acquisition Costs as well. If you pay out 500 USD in affiliate fees, the numerator of you CAC calculation increases by 500. Paypal used to pay each user who referred a new user 10 USD. The new user got 10 USD as well. This strategy increased the CAC by 20 USD.

Links

• How To (Actually) Calculate CAC by Andrew Chen and Brian Balfour
• Formula To Calculate & Reduce Customer Acquisition Cost (CAC) by Profitwell.com
• How Is CAC Changing Over Time? by Profitwell.com
• Your Average CAC is Lying to You — What to do Instead by Brian Balfour
• The Basics of Growth — User Acquisition by a16z

If you are looking for the Swiss army knife of startup metrics Customer Lifetime Value (LTV, sometimes also CLV or CLTV) is your best bet. It tells you the average total profit that one of your customers brings in. In other words: Customer Lifetime Value tells you how much profit a customer will (statistically) generate over the entirety of your business relationship.

Combined with your Customer Acquisition Costs (CAC), the LTV gives you the “holy grail” of startup metrics: The LTV/CAC ratio. It’s commonly recommended to aim for LTV/CAC ratio of 3, i.e. each customer should on average bring in 3 times as much profit as it has cost to acquire them.

The beauty of the LTV metric is that it combines the financial and retention characteristics of your business.

Some metrics like the Average Revenue Per User (ARPU) inform you about the average revenue each user spends in a given time period. But it doesn’t say anything about how long customers stick with you, i.e. for how long you can capture that revenue.

For that we have retention metrics, most notable customer and revenue retention. Customer retention is about how long customers stick with you, but it ignores how much they spend over that time period. Revenue retention is the revenue number you can retain from last period to the current one but doesn’t shed light on the number of customers or for how long you can retain them.

Sometimes improving one of these metrics can have an overall negative impact on your business success which you can’t tell by just looking at this one metric. For example, if you give generous renewal discounts you can end up in a situation where customers use your product for longer (retention yay), but spend less over the entire course of your business relationship (LTV nay).

The LTV ties both dimensions together. It can be improved by keeping customers for longer or by making them spend more over the same period of time.

Formula

In the simplest sense Customer Lifetime Value is calculated as LTV = Customer value • Average customer lifespan.

In practice the two most common ways are to either use the Average Contract Length (ACL) or the churn rate as input for how long your customers’ lifetime is. Which one you choose depends on your personal preference and especially on what data points you already have available. In the special case of negative churn (see below for details) you might prefer the ACL version.

• ARPU = Total revenue in time period / # of users in time period
• Gross margin = (Revenue - Costs of goods sold) / Revenue. Cost of goods sold for SaaS companies includes items like hosting costs, expenses for customer support, or 3rd party software.
• ACL = How long your contracts with customers are on average. See here for an exemplary calculation. See here for an exemplary calculation.

Let’s quickly define retained customers: Customers from last period which are still customers in the current period.

• Churn rate = 1 - Retention rate
  Your retention rate can be either based on customer retention or revenue retention:
  • Customer retention rate = Retained customers / Total number of customers in last period
  • Revenue retention rate = Revenue of retained customers / Total revenue in last period
    Note: Your revenue retention rate can be greater than 1. Check the What to watch out for section below for an explanation and remedies.

• As most metrics, LTV is also calculated for a given period of time. The most common periods are monthly, quarterly, and annual. You can use the billing frequency of your product to determine which option is suitable for you.
  • If you charge your customers monthly, then look at monthly ARPU and monthly churn or average contract length in months.
  • If most of your revenue comes from annual contracts, then an annual ARPU and annual churn or ACL in years is preferrable.
  • If it’s an even split between both, then you could think about calculating two separate LTVs — one on a monthly, the other one on an annual basis.

Why it matters

• The Customer Lifetime Value gives you an easy to digest insight into how healthy your business is in terms of monetizing and retaining customers.
• For more detailed and actionable insights you can split your LTV by different dimensions. This way you can easily identify your most valuable groups of users or best performing ad channels. These splits can be geographic, by acquisition channel, by marketing campaign, the pricing plan they are on, or even the sets of features they are using.
• Knowing your most — and least — valuable customer segments directly helps in making important strategic decisions such as: What acquisition channels to focus on. Dive deeper into why certain pricing plans or feature sets aren’t as valuable as you might think. It also helps you carve out your ideal customer profile (ICP).
• Segmented LTV is also one useful metric to evaluate the value of changes you made to your messaging, new features, advertising channels, marketing campaigns etc.

How to improve it

• Treat customer support and customer success as retention drivers and not as cost centers. The quicker and more effective your support can resolve the customers’ problems, the better the onboarding experience and guidance from the customer success teams, the longer your customers will stick.
  In a subscription setting, better retention directly translates into more revenue per customer. Even for infrequent of one-time purchases, happier customers who use it for longer will indirectly improve your acquisition of new customers through increased word-of-mouth or more user-generated content. Also the chance for cross-selling is higher.
• Improve your customers’ average order value. This can have an immensely positive bottom-line impact, but you have to be careful to not overdo it and alienate customers.
  • One straightforward way to increase the customer value is by increasing your prices.
  • Other common strategies are cross- and upselling. Maybe there are add-ons that you can make more attractive so more customers buy it in addition to their main product. Or you can try to shift them towards higher pricing plans, either by redefining what your plans include, by changing the messaging, or by introducing an unattractive option to nudge them towards a more expensive plan (known as the decoy effect).
• Reduce your costs to increase your margin and profit. Be careful here to not cut costs that have detrimental effects on your revenue or retention. You definitely don’t want to cut support costs if that reduces customer satisfaction and retention. You also don’t want to throw out an expensive but high-productivity software that your marketing team successfully uses to target the right leads, indirectly improving ARPU.

What to watch out for

• Most importantly: Customer Lifetime Value is calculated based on the profit customers bring in, not the revenue they produce. Using just revenue puts you at risk of overestimating the health of your business strategy. Crucially it also makes the LTV/CAC ratio meaningless. By looking only at revenue, you might still lose money with every new customer, even if your LTV/CAC ratio is greater than 1.
• Your LTV can be lower than your acquisition costs for a period of time — assuming that you have sufficient liquidity to walk through that vale of tears. At some point though you will have to get the lifetime value above the CAC. This can be done by either lowering your CAC or increasing your LTV.
• While we’re generally aiming for high LTVs, there are cases where a low customer lifetime value is sustainable as well — namely when the costs to acquire the customers (CAC) are even lower. This can for example, be the case for e-commerce shops that sell relatively cheap products and have a low average order value.
• Customer Lifetime Value is most common for subscription products and services but works almost equally well for infrequent and one-time purchases. You might have to get a little creative in defining what “lifetime” means in those cases and what your period length is. One idea could be to define lifetime as “how long do customers on average actively interact with the brand after their purchase”.
• As every average, Customer Lifetime Value get less useful the more variance you have in your data. If most of your revenue comes from a small set of customers, or there’s a very uneven distribution of how long customers use the product, you should segment your LTVs.
• The metric only becomes meaningful with a big enough sample size. If you only have a few dozen customers, then losing two or three in one month will already drastically increase your churn rate and lower your LTV. As a rule of thumb: When you ARPU and/or churn rate (or Average Contract Length) significantly differ from month to month, then you’ll also see a very “spiky” LTV graph with no clear trend over time.
• Negative revenue churn: This happens when you have positive revenue retention. Assume retained customers spend more than last period, for example, due to cross-selling, up-selling, or usage-based pricing. If this increase overcompensates the revenue you lose due to churn, the revenue retention rate will be > 1. Since the revenue churn rate is 1 - revenue retention rate, your revenue churn will be negative, and the LTV formulas above becomes invalid (your LTV would be negative).
  • The easiest way to get around this is not using churn but the Average Contract Value for calculating your customer lifetime. Of course you have to ensure that you calculate the LTV for every period using the same formula.
  • If you want to stick with the churn approach, there are some complex solutions described in this article and in this post. Most startups don’t constantly achieve positive revenue retention, though. If it’s just an outlier for one or two periods these calculations seem like overengineering. Though not formally correct I would just take the LTV of the previous month and increase it by 5%-15%. If your revenue retention rate is consistently above 1 and you have to use the churn rate, then it can make sense to try one of the two solution.
• A quick note on the difference between ARPU and LTV: ARPU = how much revenue you get (on average) from one customer for some time period (e.g. per one month or per one year). LTV = how much revenue you get (on average) from one customer over the sum of periods they stay with you.

Links

• Guide to Customer LTV: Calculation, Formula & SaaS Business Strategies by Profitwell.com
• What’s your TRUE customer lifetime value (LTV)? by David Skok
• SaaS Metrics: Negative Churn by Ferenc Huszár

The concept of payback period is pretty simple: You will incur some costs to acquire a customer, i.e. your Customer Acquisition Costs (CAC). Each customer will also generate revenue and a profit for you. The payback period shows how long it takes until the customer has paid back his acquisition costs.

It’s an important metric to properly manage your liquidity. Also, it helps put the Customer Acquisition Costs into perspective: Assume you have a CAC of 20 USD — is that a lot? Is it perfectly fine? Knowing how long it takes helps you put this absolute number into perspective.

The payback period is one key metric to understand the efficiency of your acquisition model.

Formula

• Customer Acquisition Costs: The average costs to acquire a customer. For details see Customer Acquisition Costs (CAC).
• Average profit per period = Average revenue per user (ARPU) in this period * Gross margin in percent.
• If you sell monthly subscriptions, then you have to look at monthly intervals. For annual products, you’re looking at one-year periods.
• Example: Your CAC are 80 USD and you charge 25 USD per month. Your gross margin is 75%. Then you payback period is 80 / (25 • 0.75) = 4,2 months.

Why it matters

• Knowing your payback period helps you avoid liquidity issues. If you're getting your acquisition costs for each customer back after 10 months you will need significantly more cash on the bank than if it only takes 4 months. Don’t know your payback period and chances are you’ll wake up one day seeing your bank account lighting up red.

How to improve it

• Bring your Customer Acquisition Costs down. Find more efficient acquisition channels, make your ads more relevant, improve your messaging. A more long-term approach is to invest in content marketing and brand building now to reap the benefits of more inbound leads in the future.
• Increase your revenue per user. You can raise your prices, up- and cross-sell your customers, make the more expensive plans more attractive, try to expand team sizes for enterprise customers, or add some usage-based pricing components to your plans (assuming that customers love and heavily use your product…).
• Improve your gross margin by cutting hosting costs, or identifying some software tools that you’re paying for but don’t really need anymore.

What to watch out for

• The universe you base your payback period calculations has to be the same one you used for calculating the acquisition costs. Assume you calculated the CAC for just enterprise customers.
  They are on annual contracts, although you also offer monthly plans for individual customers. Then you have to calculate your payback period on a yearly basis with just the average profit these enterprise customers bring in. Otherwise, you’re comparing apples and oranges.
• Similar to the CAC it usually doesn’t make much sense to compute just one payback period for the entire business. If your customer base is very homogeneous in terms of contributed revenue per month/year, acquisition channels and costs, and need for support just one payback period can suffice.
  But if you have small, medium, and enterprise plans, maybe special add-ons for enterprise customers, and you users come in through very different channels, then you have to segment these groups and get group-specific payback periods.
• By definition, payback periods only exist for subscription-based business models or at least those that feature predictable repeat purchases. In the ecommerce world, and often also in the consulting or freelancing sphere you will need to recover your CAC with just one transaction. Hence the payback period here would be 1.
• Payback period does not recognize churn. It tells you how long you have to keep an average customer to recoup your CAC. But this does not automatically mean that they do stick around that long. It can be the case that your payback period for some customer groups is 7 months, but the average customer retention is only 5.5 months. Then you have to either significantly lower your payback period for this user group or improve retention.
• You have to calculate it based on the profit your customers bring in, not their revenue. If you have high variable costs or costs to serve a given customer (i.e. low margin), then your payback period will be much longer given the same revenue and CAC than for high-margin businesses.
• In a SaaS context you should not use extrapolate the graph of your payback period and use it as long-term revenue projection. The longer customers stick with you, the more likely they are to change the pricing plan they are on. Up- and downgrades are only partially included in the payback period, though.

Links

• Examining the Sale Efficiency in Public SaaS Companies by Alex Clayton
• How to calculate and reduce payback period by paddle.com