One of the mysteries I hoped to solve when I embarked upon this little SaaS metrics mathematical journey was the reality behind “The SaaS Company Magic Number” introduced by then Omniture CEO, Josh James, and immemorialized by my pal Lars Leckie over at Hummer Winblad. In principal, this number tells you how aggressively you should be spending to build up your customer base: “the key insight is that if you are below 0.75 then step back and look at your business, if you are above 0.75 then start pouring on the gas for growth because your business is primed to leverage spend into growth. If you are anywhere above 1.5 call me immediately.”-Lars Leckie’s Blog
In this installment of the SaaS metrics series, I will show why this benchmark works, and introduce what IMHO is the single most important SaaS financial metric for measuring the overall health of a SaaS business. Now, being as Josh and Lars have already laid claim to “The” SaaS Magic Number, I really have no alternative but to put all humility aside and dub my latecomer SaaS metric as “Joel’s” SaaS Magic Number (After all, I did similarly snag “The” Top Ten Do’s and Don’ts of SaaS, so it’s all good.)
SaaS Metrics Rule of Thumb #9 – Joel’s SaaS Magic Number
The truth be told I feel a bit guilty about the name, because Joel’s SaaS Magic Number is neither magic, nor mine really. It’s simply the the average customer rate of return, or rather the inverse of the average customer baseline break-even, 1/BE0, that has so consistently popped up as a driver and a constraint in the SaaS metrics rules-of-thumb.
| J | = | ARR – ACS |
| CAC |
| Joel’s SaaS Magic Number | = | Average Customer Rate of Return |
Where “ARR” is the average recurring revenue per customer, “ACS” is the average recurring cost of service per customer, and “CAC” is the average customer acquisition cost. Customer rate of return is powerful, because it measures the economics that make a SaaS business work (or not), whereas the individual revenue and cost metrics are simply accounting figures that in isolation say little about the health of the business. Let’s recap some of the things we know about this nifty magic number from earlier SaaS metrics rules-of-thumb.
| Joel’s SaaS Magic Number Rules-of-Thumb | ||
|---|---|---|
| J | [ ARR – ACS ] ÷ CAC | average SaaS customer rate of return |
| 1/J | BE0 | best case SaaS company time to profit |
| limiting | g = a = J | maximum, profitable rate of growth g or churn a |
| approaching | g ? J or a ? J | dramatically delays SaaS time to profit |
| exceeding | g > J or a > J | SaaS company will never be profitable |
| increasing | ?J by ?ARR or ?TCS | upselling & lower TCS accelerate profitability |
| recommended | J > max( g , a ) | or a + net growth (g – a) |
| benchmark | J > 50% | per year is generally very healthy |
Intuitively, the long run profitability of SaaS companies requires
the recurring contribution of current customers to cover the acquisition cost of new customers,
therefore the average customer rate of return for SaaS companies
must exceed both the current customer churn rate and the new customer growth rate.
So, what is a good value for Joel’s SaaS Magic Number? Well, I don’t think anyone could argue with 1, a valiant goal, but rather optimistic. If your average customer rate of return is 100%, chances are you don’t need VC money, because you can bootstrap your way to the top given that your customers pay for themselves within the first year. Personally, I’m prejudiced toward profitable growth in the long run, not just growth (call me old fashioned). In which case, I’d recommend shooting for something in the J ≥ 50% per year range, implying a minimum time to profit of 2 years (probably more like 4 years with growth and churn) and the ability to grow annually at upwards of 50%.
| Customer rate of return is powerful, because it measures the economics that make a SaaS business work (or not), whereas the individual revenue and cost metrics are simply accounting figures that in isolation say little about the health of the business. |
At the very least, you should set a goal for average customer rate of return that significantly exceeds your both your churn and target growth rates, J > max( a , g ) = a + net positive growth (g-a), unless you are prepared to burn cash for a very, very long time and build up a very, very large accumulated loss deficit on your balance sheet. But, focusing exclusively on growth to the detriment of profit is a dangerous game in SaaS. Most SaaS markets are not as wide-open as enterprise software markets were twenty years ago. Personally, I believe there is very little a weak executive team can do for $110M in expense at a $10M loss, that a strong executive team can’t do better for $90M in expense at a $10M profit. To each his own.
I’ll let the interested reader refer to Lars’ original blog post for the specifics, but after a little comparison I’ve concluded that “The” SaaS Company Magic Number is roughly equivalent to following ratio in my SaaS metrics model.
The SaaS Company Magic Number = ARR / CAC
(for newly acquired customers with ARR measured annually)
Thus, the difference between Joel’s SaaS Magic Number and The SaaS Magic Number is the explicit inclusion of recurring cost of service, giving IMHO a stronger metric for the financial health of SaaS companies that includes the full measure of both recurring revenue and total cost of service. In relation to The SaaS Magic Number we have the following:
Joel’s SaaS Magic Number = ( [ ARR – ACS ] / ARR ) x The SaaS Magic Number
Joel’s SaaS Magic Number = Contribution Margin x The SaaS Magic Number
Let’s plug in some numbers and see what we get. If we want a quick and dirty comparison of The SaaS Magic Number to Joel’s SaaS Magic Number, then the easiest approach is to assume a contribution margin of 50%. This isn’t so outrageous given the cost structure of many SaaS/software companies that even in the early stages tend to spend about 50% on sales and marketing (CAC) and 50% on everything else (ACS).
When Contribution Margin = 50%
Joel’s SaaS Magic Number = The SaaS Magic Number / 2
Aha! Now we have a rhyme to the reason of why Lars just might want to talk to you if The SaaS Magic Number is > 1.5, because you are primed for a profitable, annual viral growth rate of up to 75%, something VCs like very much. However, I have to take issue with “pouring on the gas” if your The SaaS Magic Number is simply above .75 . In this example, your average customer rate of return is about 38%., giving you a best case time to profit of 2.7 years. In reality, if your annual growth is approaching 38% (which your investors will very much desire) and your contribution margin is anything less than 50% (very likely in the early stages), then your real time to profit is probably double or triple that (8 years!), because you aren’t even covering your acquisition costs let alone your total cost of service. Depending on your actual contribution margin, you may very well be burning through wads of cash by stepping on the gas. Personally, I’d up this requirement to a clean The SaaS Magic Number of > 1.0 with the added requirement that ARR ≥ 2 x ACS, so that your contribution margin can quickly recover acquisition costs. This is consistent with the prevalent SaaS metrics conventional wisdom that you’d like to recover your customer acquisition cost in the first 12 months, but with the added requirement that you need to pay for your ongoing service costs as well.
Almost done! There is one more post in this SaaS metrics series, because clearly we can’t stop until we reach SaaS Metrics Rule of Thumb #10 and have a complete SaaS metrics top ten list. The next and final post in the SaaS metrics series entitled SaaS Customer Lifetime Value Drives SaaS Company Value will take things up a level to examine the highest level SaaS financial metric of company valuation, and it has some fun, no-holds-barred extra credit math notes for the math heads. Stay tuned!
SaaS Metrics Math Notes: What’s YOUR SaaS Average Customer Rate of Return?
The theory is only useful if we put it into practice, so today’s SaaS Metrics Math Note includes the homework assignment to calculate the average customer rate of return for your SaaS business (your very own SaaS magic number!). In practice, calculating average customer rate of return requires a few decisions about how you’re going to measure recurring revenue and costs, but the basic formula for calculation is as follows.
| [ Total Recurring Revenue – Total Recurring Costs ] ÷ Total Number of Customers |
| Total Acquisition Costs ÷ Number of New Customers |
Total Acquisition Costs should include all sales, marketing and overhead expenses directed at new customer acquisition within a given period, and Total Recurring Costs should include just about everything that is left over in the same period, including everything that enables you to service your customers from rent to support staff, even accounting because you gotta send out the bill. You can dance around the idea of fixed vs. variable costs, but in most SaaS companies pretty much all costs are variable in a roughly 3+ month time-frame. Also, note that the average customer rate of return time-frame will be the same one that you choose for your recurring costs, i.e., percent per month/quarter/year respectively.
I recommend making two charts, one with the average customer rate of return (you can also include the other two relevant SaaS metrics of growth and churn for comparison) and a second with the component values of ARR, ACS and CAC. This example highlights the positive impact of both lowering total cost of service (decreasing ACS/CAC) and upselling (increasing ARR) on average customer rate of return as implied in SaaS Metrics Rules-of-Thumb 6, 7 & 8.
Whether you choose GAAP, cash or any number of other variations as your SaaS metrics basis for calculation is probably fine as long as you are consistent throughout. Personally, I prefer cash.
Plus, you’ll want to pick a measurement period that will highlight the long term trend over short term random fluctuation, because these profitability and growth SaaS metrics are mid-to-long term measures. One month is probably too short and one year is probably too long. So, some sort of moving average around 3-6 months is probably best for most SaaS companies. You can also get fancy with things like customer segmentation to see how the average customer recurring rate of return varies by segment. Consider that extra credit!
Check out the rest of the SaaS Metrics Rules-of-Thumb
- SaaS Metrics Rule-of-Thumb #1 – SaaS Churn Kills SaaS Company Growth
- SaaS Metrics Rule-of-Thumb #2 – New Customer Acquisition Growth Must Outpace Churn
- SaaS Metrics Rule-of-Thumb #3 – Viral Growth Trumps SaaS Churn
- SaaS Metrics Rule-of-Thumb #4 – Company Time to Profit Follows Customer Break-Even
- SaaS Metrics Rule-of-Thumb #5 – Best Case Time to Profit is Simple Break-Even
- SaaS Metrics Rule-of-Thumb #6 – Growth Creates Pressure to Reduce Total Cost of Service
- SaaS Metrics Rule-of-Thumb #7 – Churn Creates Pressure to Reduce Total Cost of Service
- SaaS Metrics Rule-of-Thumb #8 – Upselling and Upgrades Accelerate SaaS Profitability
- SaaS Metrics Rule-of-Thumb #9 – Joel’s SaaS Magic Number
- SaaS Metrics Rule-of-Thumb #10 – SaaS Customer Lifetime Value Drives SaaS Company Value
Really helpful Joel! These SaaS metrics rules-of-thumb that you took time writing and elaborating for it to be shared online is very much appreciated! i go crazy over one and here you have provided 10 rules-of-thumb! One would definitely find one that suits best their SaaS platform, or better yet use all of those!
[…] Read every post on SaaStr and blogs like David Skok’s, Mark Suster’s and Joel York’s. […]
[…] metrics, more advance metrics such as the customer lifetime value, churn cohorts and the various SaaS magic numbers may also be […]
[…] metrics, more advance metrics such as the customer lifetime value, churn cohorts and the various SaaS magic numbers may also be […]
[…] is hard to fault us for not investing more in marketing because we clearly had not solidified the Magic Number math that is essential for justifying increases in marketing […]
[…] is hard to fault us for not investing more in marketing because we clearly had not solidified the Magic Number math that is essential for justifying increases in marketing […]
Is there a book you could recommend to help me understand these concepts.
I am trying to figure out what products/services we should focus on selling and how fast we can grow so our MRR will meet or exceed our expenses
Joel
Just discovered your blog. I have been working with a client who has a terrific SaaS product with a couple of major customers. The last two years have been constant development and now I am searching for information on different organization struectures and business models which have been successful in the SaaS world.
Ray
[…] Joel York takes this number a bit further by including Average Cost of Service (ACS) into the formula with the conclusion that the cost of service should be covered when calculating the Average Customer Rate of Return for the SaaS business. According to York, the Customer Rate of Return is the most powerful metric that a SaaS business can be measured on as it really shows whether the SaaS business will be a business or not. Therefore, York concludes this number (J) to be calculated in following way: […]
Hi Joel
I cannot get the calculation to work for the life of me in a way that makes sense. It seems you gave a few variations in how to calculate the average customer rate of return and when I put it into my numbers, they all vary in the output.
1st Formula: (ARR – ACS)/CAC
2nd Formula: (ARR – ACS)/ ARR x (ARR/CAC) where (ARR-ACS)/ARR = Contribition Margin and (ARR/CAC)= The Saas Magic Number
3rd Formula: (ARR-ACS)/# of Customers)/(CAC/# of New Customers)
Am I missing something?
[…] Joel’s point of view on Cost to acquire a customer is illustrated here at Chaotic-Flow.com Cloudonomics 101 – Creating a Financial Plan for your SaaS or Cloud Computing Business View […]
Hi Alex,
It is difficult to give a good answer without seeing the actual calculations, but my first impressions are the following.
1) I would not adjust for network effects, these rightfully increase J (customer rate of return), because you are spending less to acquire customers when you have strong word of mouth. This is good, and it should show up as a higher J. I usually handle wide variations simply by using a moving average for CAC and ACS. ARR is generally very stable.
2) 5 seems very high…so to confirm the calculation formula for J
a) are you including all sales/marketing staff costs in CAC?
b) are you including all product/service/admin costs in ACS?
c) are you using new customers to calculate CAC,
but total customers to calculate ARR and ACS?
d) are ARR and ACS measured in annual recurring costs?
(this should be the case even if your measurement period is semester)
3) It seems unusual that CAC should vary so dramatically…what is the source of variation?
a) are out of pocket acquisition costs larger than staffing costs?
b) are the number of new customers small, e.g., 1 then 5, then 2, etc.
4) I would first recommend coming up with formulas (probably 6 month moving averages) that accurately represent stable values for average CAC, ARR and ACS separately. Then you can be sure your J is correct…even if it is 5!
…if you want to email me a spreadsheet to jyork [at] chaotic-flow.com, I’d be happy to take a look and email you back better answers confidentially.
Regards,
Joel
Hi Joel,
first thanks for your good work on chaotic-flow!
While computing CAC RATIO/j related metrics for my own B2B SaaS company I two points that appear to be interesting:
First:
Sometimes, during periods of low investment in CAC, we still had grow.. I understand that most of this growth should be attributed to mouth-to-mouth/network effects not our direct sales/mktg efforts or past CACs (something that I’ll comment later below).
As we had growth but with a low CAC, our J number was very high (5, for example) during this period, giving misleading information at a first glance (a very appealing ROI)
Is there, on your vision, a good way to adjust to this fact?
I thought on an adjusted version that discounts the network effects:
original J = (ARR(t) – ACS(t))/CAC(t-1)
adjusted J = (ARR(t) – ACS(t) – a(ARR(t-1)-ACS(t-1) ) )/CAC(t-1)
where a is the supposed growth due to network effects (10%, for example)
Second:
We work with semesterly metrics as they are fit better our sales cycles. Most of the leads are converted or not during this 6-months period. But sometimes (I would say 30% of the cases), we see old prospects returning and closing the deal without more efforts! Something that is not attributed to CAC(t-1), but to CAC(t-2) or even CAC(t-3)
What do you think on this two cases?
Alex
Hi there Joel, I was just doing some research when i came across your site. very good resource i will add. your wife is right about too much theorizing!. anyway, just had a question is there actually any Saas business making money!. or is the idea behind saas not practical in the real world (due to human factor), though in principle it may be sound on paper and in theory but doesnt stand up in real practical world.