Why are metrics revenue-focussed?
Metrics involve numbers. They do what they’re meant to do. They replace complexity, and tell us how well we’re faring in a given direction.
Whether the complexity is masked to lie to ourselves, or to side with a direction of ruthless growth above all else, a metric still does what it’s meant to do.
And they are constantly worked upon, sometimes, even made up, almost always to increase revenues and profitability. That is why we come across questions such as “How do you improve MRR for more revenues?”. Or “How do you reduce churn to increase revenues?”. Or “How do you increase the customer LTV to improve profits?”
Groupon thought CSOI was a smart method to report profitability, ahead of its IPO, that drew attention from marketing and acquisition-related expenses (non-cash and otherwise), and requested the investors to look at the profits instead, and got into trouble with the SEC.
Small wonder, then, that most revenue related metrics can fail us. Unless, of course, they’re tied back to real value that a business seeks to deliver.
As Seth Godin, so eloquently puts it, “A useful metric is both accurate (in that it measures what it says it measures) and aligned with your goals. Making your numbers go up is pointless if the numbers aren't related to why you went to work this morning.”
When thought well, ARPU does help put the two pieces - revenue and value - together.
How do you optimize for ARPU?
Conventionally, ARPU was measured based on the initial contract value during sign up. This made sense, when annual contracts were a bigger source of revenues. Now, businesses find themselves more profitable to raise ARPUs through cross-selling, up-selling or upgrades than increasing customer base at higher customer acquisition cost.
Upgrades, Upsells and Cross-sells
Here’s an example.
Adobe announced its quarterly financial results in March 2017. A record quarterly revenue of $1.68 billion in the first quarter of fiscal year 2017 was driven predominantly by the Creative Cloud subscriptions.
Adobe Creative Cloud has been around for 3 years, and the annualised recurring revenue of Adobe Creative Cloud is pegged at $3.76 billion. As of today, it reports over 9 million subscribers. The numbers are staggering.
What does this have to do with ARPU?
The Street reports, “On the call, Adobe mentioned the pending expiration of initial 3-year Creative Cloud agreements with many enterprises could boost its average revenue per user (ARPU), since many of the new deals could involve the entire suite”.
Treading lightly in terms of reviewing the company’s numbers and its competitive positioning, this growth was largely driven by subscription adoption and retention of Adobe’s cloud-based subscription products.
Creative Cloud’s growth is much more multi-faceted. Many users moved from the freemium pricing to paid plans, as a result of which, there was a dramatic payoff in ARPU.
Clearly, a good ARPU indicates good revenue margins. And in Adobe’s case, it was the significant upgrades from free to paid plans that helped its cause.
But, what if this healthy cohort of freemium to paid users are exhausted within the year, or within the quarter? How do they maintain the same ARPU in the subsequent years, without burning too much in Customer Acquisition Cost (CAC)?
Play around with Pricing and Segmentation
Let us take the example of Statuspage.io, a status and incident communication solution. Statuspage.io’s pricing page indicates transparency. In their blog post about how they increased their ARPU by 2.4x.
In the post, Steve Klein, Co-Founder of Statuspage.io says that raising the highest listed pricing to 8x of the first pricing iteration, and adding deeper pricing segmentation, helped bolster the ARPU. He went on to add that prioritizing feature development in the product and aligning that prioritization with the MRR growth helped in building a great product, that people wanted and were willing to pay for.
And finally, he adds that bumping the Enterprise pricing to 20x the base price helped them understand how enterprise customers looked for specific features, that did not really belong to the core product. Some of these include features such as single sign on for security, SLAs, escalation support to the highest level, access control etc.
This is to say that, segmenting customers based on the plans and pricing, prioritizing and developing specific feature sets that can be nonlinear in terms of value, and focusing on enterprise readiness, tend to have a huge bearing on the ARPU.
It also helps you understand which target segment find value on what you offer, so that you can double down your marketing and sales activities to win that segment.
Say, you are selling a SaaS product, Meet that allows you to schedule, conduct, manage and record virtual meetings. Meet has three different pricing plans, viz. Basic at $30, Pro at $300 and Enterprise at $3000. Meet caters to SaaS-based customers from various stages of growth and sizes including 1000 customers which are startups, 50 SMB’s and 10 Hyper Growth/Enterprise stage businesses.
The customers across the different plans are segmented as below.
The total ARPU calculated would then be calculated as:
(((985+15+0)*30)+((15+45+0)*300)+((0+0+10)*3000))/(1000+50+10) = $69.34.
This indicates that you can focus on growth avenues for the Basic plan targeting the customer base of Startups and SMBs.
However, if you segment the ARPU depending on the customer segment in a similar fashion, you will see that the ARPU for Hyper Growth and Enterprise segment of customers is high.
Which means, that Meet can consider building a few enterprise-ready features that will appeal to this segment. This can maintain the health of the ARPU consistently.
This, of course, comes with a caveat. Because, what is life, without a few occasional buts?
But can ARPU be viewed independently? Are ‘Averages’ a valid metric?
As mentioned in the beginning of the post, there are some who believe ARPU is not a value metric but a vanity metric. Why?
The post points to the Power Law of Distribution, that goes to say a small change in one quantity can result in a proportional relative change in another, and it usually appears when there is a significant variance in data. This is especially true when money is involved.
In our example, in terms of the pricing plans, there is a significant difference between the Basic and Enterprise plan. So, the ARPU is bound to increase proportionally, if we add even one additional enterprise customer. This could skew the average.
While the conclusion of the post is slightly misleading, it also sensibly points out that it does not make sense to optimize the average alone, but instead look at the bigger picture.
How ARPU operates is entirely contextual, and it can become a vanity metric, if not tracked in the context of Net MRR Growth Rate, CAC, LTV:CAC Ratio, Average Margins Per User (AMPU). The mean or average is a highly useful and informative value, but not to be used or optimized alone.
ARPU has a significant impact on LTV. For SaaS companies, LTV can be accurately calculated taking into consideration the gross margin percentage as below:
LTV ($) = (ARPU ($) * Gross Margin (%)) / Monthly Customer Churn ($)
Alternatively, it can be calculated based on the customer lifetime as below:
LTV ($) = ARPU ($) * Customer lifetime (no. of months)
Brian Balfour, Founder of ReForge, in his talk at SaaSFest highlights that (high-price-high-touch) businesses with high ARPU and CAC operate differently from those (low-price-low-touch) with low ARPU and CAC. He adds that you need to be strategic about which models you use with your channels’.
Accordingly, here is an archetype of Meet’s ARPU, distribution channels, CAC etc.
There is no industry standards for a good or perfect ARPU, as such. It would make more sense to judge a good ARPU based on the growth percentage from the previous month or year.