Answer: Net Retention

Of course, most people in the investment community speak in terms of ARR(Annual recurring revenue) multiples. Or till recently EBIDTA multiples. That works great, for the most part. However, there are always cases where there is a huge difference in valuation even in SaaS companies that have the same ARR. Typically this number lies in 7x-10x ARR multiples. For some lucky(?) companies (erm, Snowflake) this can get as high as ~100x ARR multiple. Why this huge difference?

The answer to the above question is the lynchpin of the SaaS business model. In SaaS, more often than not, it costs more to get one customer than how much money we get from the customer in year 1. We lose money on customers when we acquire them. The metric which accurately captures this concept is CAC (Customer Acquisition Cost) Ratio.

The above metric CAC Ratio is typically < 1(according to one benchmark, the average CAC Ratio is 0.71). The inverse of this metric (1/CAC Ratio) tells us how long does it take to get the money we spent to acquire a customer from the customers as our revenue. Well, the good thing is we don't just have 1 year to recoup our losses. SaaS is a recurring business, so only measuring the year 1 revenue does not tell the whole story.

For starters, the money we get from the customer, LTV (Customer lifetime value) should be the total amount we get across the lifetime of the customer. In our simple model, the money we get from customers can be modeled as below:

Customer lifetime means how many years (or months) a customer stays with us. To calculate it there are 2 ways.

1. Historical measurement — this is not reliable because the external and internal conditions are always changing.

2. Predictive measurement — To predict the future value, we can look at churn today. Churn is easier to measure as well.

If we modify the above equation considering all the customers, the Revenue changes to Average Revenue per customer (ARPA). Let us also take into consideration the predictive measurement, the equation now changes to:

Churn is really, really important when we measure the value of a company. Churn impacts the LTV and hence the future revenues. Look at the below chart to see it for yourself:

Churn is really, really important when we measure the value of a company

Another way to increase the value of the company is to increase the average revenue per customer (ARPA). One way to do that is to hunt bigger customers. But hunting is hard and we might not be able to impact it directly in short term. The second way is to make sure that customers spend more in subsequent periods after we land them. i.e. Upsell and Cross-Sell.

Any metric that directly predicts the value that a company delivers should incorporate the growth and the churn aspects. Hence the Net Retention metrics. Let's first look at the metric that's all rage now, Net Dollar Retention (NDR):

NDR is a really simple metric to calculate and absolutely captures how the business will change in the future. Direct dollar number articulation really helps put a sense of urgency since we are no longer using any abstractions — it's pure, raw dollars.

If we want to compare a metric across companies, we use the second Net Retention metric called Net Revenue Retention (NRR) Rate.

ARR growth might be the form, but NRR is the class and class is permanent

There is of course one big elephant in the room I ignored when I made this assertion around Net Retention. That is ‘new ARR growth’. New ARR growth absolutely impacts the valuation. But the thing with growth is — it eventually slows down as the base keeps on getting bigger. Over the long term, the driving factor for valuation changes from growth to NRR and profitability (Gross Margin). Using a cricket metaphor during IPL time, ARR growth might be the form, but NRR is the class and class is permanent. Considering that profitability is something that companies can control to an extent relatively, I’d conclude that Net retention is the most important metric for SaaS companies.

Note: All the metrics are simplistic representations. There are more nuances (RR at end of the period vs the beginning of the period, Annual periodicity, etc.) if you want to use these for any accounting/valuations.