Over the past few years, I’ve advised a large number of CEOs and CFOs of software-as-a-service (SaaS) companies all of whom shared one common challenge: which of the many SaaS metrics frequently cited they should track. I’m frequently asked, “I just need to know a few key metrics that can tell me how I’m doing and that I can report and trend on a consistent basis.”
There are four key questions that I begin the analysis of any SaaS business by asking; I consider the four answers as fundamental metrics.
These four key questions are:
- How much does it cost you to acquire a customer?
- How “big” is each of those customers?
- How efficiently do you delivery your offering to your customers?
- How effectively do you retain your customers?
1. Customer Acquisition Cost (CAC)
How much does it cost you to acquire a customer?
There are quite a few approaches to calculating CAC, from those that treat it as a ratio (dollar spent on sales & marketing to dollar of new sales) to those that account for gross margin to every variation in between.
For the purposes of this series, I’m going to use a relatively simplistic definition: on average, how much must be spent on sales, marketing, and setup services to acquire and on-board a customer?
CAC = sales & marketing expenses + setup expenses - setup revenue
number of customers acquired
Example
For a particular time period (month, quarter, year, etc.):
- total marketing expenses were $150K
- total sales expenses were $250K
- total setup/implementation expenses were $75K
- setup/implementation sales/bookings were $50K
- 10 new customers were acquired
CAC = $150K + $250K + 75K - $50K = $85K per customer
5 customers
Over this time period, our average customer cost $85K to acquire.
If you want to be a little more sophisticated, you can try to account for time-shift issues (marketing spend in Q1 likely impacts sales/bookings primarily in Q2, etc.) by measuring expenses for prior periods (e.g., marketing expenses in Q1, sales expenses in Q2, but setup/implementation expenses in Q3).
2. Annual Revenue per Customer (ARC)
How big is each of those customers?
This metric is quite straightforward with the only variable being over what time period to calculate it. It’s important to note that his metric is only concerned with the annually-recurring revenue of a customer and does not include on-time revenue due to setup/implementation.
ARC = annualized revenue run rate of customers acquired
number of customers acquired
Example
For a particular time period (month, quarter, year, etc.):
- 5 customers were acquired
- their aggregate annualized revenue run rate is $575K
*we expect to recognize $575K annually from this cohort of customers
ARC = $575K = $115K per customer
5 cust.
Over this time period, our average customer was $115K in annual revenue.
3. Recurring Gross Margin (RGM)
How efficiently do you deliver your offering to your customers?
Gross margin is a very common financial metric that is covered under Generally-Accepted Accounting Principles (GAAP). Where its calculation becomes interesting and frequently debated is 1) as it pertains to what’s included in “cost of goods sold” (COGS, another GAAP term) and 2) as it pertains to what revenue should be counted. My philosophy here is simple: to use a definition that is in keeping with what we’re trying to understand.
As it pertains to COGS, we’re trying to understand the efficiency of our business’ service delivery. For that reason, I recommend including in COGS any expenses that are needed in order to continue delivery of the customer contract even if they’re not typically counted (e.g., fractional cost of engineers in the R&D organization who may be supporting customer issues on an on-going basis). Examples of expenses that belong in COGS are fees paid to infrastructure-as-a-service provider, payroll for staff managing the production environment, payroll for staff supporting the customer, travel expenses associated with servicing the customer relationship, etc. Note, however, that this would also imply I recommend that one-time setup expenses, which are typically counted in COGS, not be counted for the purposes of this metric since they’re not recurring.
As it pertains to revenue, we’re trying to understand what this efficiency will look like over time and not just at the beginning. For that reason, I recommend calculating “recurring” gross margin separately and only including recurring revenue in the calculation.
RGM = recurring revenue - recurring COGS
recurring revenue
Example
For a particular time period (month, quarter, year, etc.):
- recurring revenue was $550K
- recurring COGS were $75K
RGM = $550K - $75K = 86.3%
$550K
Over this time period, our delivery efficiency, as measured by RGM, was 86.3%.
4. Annual Customer Retention (ACR)
How effectively do you retain your customers?
Arguably the most impactful metric in the long-term, annual customer retention (ACR) measures how effectively the business is able to retain its customer base. Interestingly, this metric is very frequently miscalculated (e.g., not accounting for new customers acquired) or obfuscated (e.g., calculating it based on revenue and including organic revenue growth within the install base).
I prefer to keep it very simple.
ACR = total customers at period end - # new customers acquired
total customers at period start
*annualize the result above if the period was a month or quarter
Example
For a particular time period (in this case, a quarter):
- 152 customers existed at quarter end
- 5 new customers were acquired in the quarter
- 150 customers existed at quarter start
ACR = 152 - 5 = 98% quarterly 150 ACR (annualized) = 98%x98%x98%x98% = 92%
Over this quarter-long time period, our customer retention effectiveness, as measured by ACR, was 92%.
This is wonderful-Thank you Bassam
On Tue, Apr 26, 2016 at 12:54 PM, Against the Herd wrote:
> Bassam Salem posted: ” Over the pas few years, I’ve advised a large number > of CEOs and CFOs of software-as-a-service (SaaS) companies all of whom > shared one common challenge: which of the many SaaS metrics frequently > cited they should track. I’m frequently asked, “I just need” >