ARR in a Consumption-based Pricing Model

Annual recurring revenue (ARR) is a vital business health metric, particularly for SaaS companies. Used often by subscription-based businesses, ARR helps CFOs forecast business growth, track and improve the performance of their sales organization, and it is often a key reporting metric (internally and externally) leading up to and after an IPO. 

However, measuring ARR is more complex for the large swath of businesses that have consumption or hybrid (subscription + consumption) pricing models. CFOs are tasked with forecasting annual revenue based on unpredictable usage and often have to report those forecasts to their board or prospective investors. Further, they also face challenges on how best to use ARR as a metric for compensating their sales team. 

To better understand some of the nuances of ARR, we convened the CFO|Circle for a conversation led by finance leaders with a lot of scar tissue in this area: Matt Peterson (SVP, Finance) and Karan Batra (Senior Director, Sales Compensation) from Attentive and formerly of Fastly (both consumption-based businesses); Matt Heist (VP, Finance & Business Operations) and Eugene Berson (Senior Director, Head of Global Sales & Success Operations) from Slack, one of the earliest usage-based companies; and Ben Thiesen (Partner) from EY who has coached several companies through the process of analyzing ARR and other key business metrics. 

Here are some of the key takeaways from the conversation:

Why is ARR Important?

The first question on CFOs minds was: if you have a consumption-based pricing model, why measure and report ARR at all? Attentive’s Matt Peterson pointed out that ARR isn’t committed in a pure consumption model, and usage is often highly variable. Many companies will measure monthly recurring revenue (MRR) and multiply it by twelve months; however, this approach won’t factor in things like seasonality or shifting customer needs. Therefore, it’s important for CFOs to decide whether ARR is the appropriate health metric for their consumption model and how it informs larger business decisions.

Of course, that might not stop your CEO, board, or investors from asking to see ARR projections, said Slack’s Matt Heist. ARR can often be a “comfort metric” for investors, but it’s important to set expectations on any limitations on measuring attrition risk, seasonality, or “land and expand” opportunities. Still, he has found that ARR can still be very useful on the sales side at predicting incoming billings and then measuring against what falls into revenue. 

ARR can also be an essential part of IPO readiness for companies transitioning to the public markets, said EY’s Ben Thiesen. Companies may choose to disclose ARR and other KPIs like LTV or CAC in their S-1 to help tell a growth story to initial investors. He cautioned, “By definition, these are all unaudited metrics, so there is typically a heavier lift internally than in areas like GAAP revenue.”

“Getting ahead with 4, 8, or 12 quarters of data supporting customer-level confidence underneath ARR and retention rate trends is so critical. Even if you’re not reporting out ARR, you likely will need to show a strong net expansion rate based on annual customer spend.” – Ben Thiesen, EY

Measuring ARR in a Consumption Model

There are a lot of different methods for calculating ARR, but starting with a clear definition of what is underlying your ARR metric is critical, said Slack’s Eugene Berson.

“ARR is about defining what drives value for your customer. The atomic unit of activity that you’re charging for — whether it’s data consumption or active usage over a given period — not only influences how you define ARR but also how you price and invoice customers.” – Eugene Berson, Slack

Ben agreed with Eugene adding that, at its simplest definition, ARR is a measurement of ongoing customer usage. To arrive at that measurement of customer usage, start by defining:

1. Who is your customer? Are you measuring usage for a single company, or are there multiple buying entities within the organization? This is particularly relevant in a “land and expand” model, where the parent company may often be the best overall measure of recurring revenue.
2. What is your basis of measurement? Is your ARR metric based on bookings, billings, or GAAP recurring revenue?
3. What period are you analyzing? Are you measuring monthly usage? Quarterly usage? There is typically a lot of variance by company, and the suitable time period also depends on a CFO’s comfortability with volatility versus predictability.
   

Leveraging Cohort Data to Improve Your Model

The lifeblood of your ARR measurement is customer usage data. Being able to segment usage across different customer cohorts can give you a more accurate forecast, said Matt Peterson.

“We have multiple tiers within our customer-base and we know, on average, what each tier will spend in a given year based on that cohort data. Therefore, when we’re looking at traffic or new customers, having that very detailed cohort data is important for predicting what tier those users will fall into.” – Matt Peterson, Attentive

Matt added that your cohort usage data ultimately aligns with actual usage and billing. “That’s a lot of data to analyze, and it’s not something finance can solve; it’s a data science problem,” said Matt. That is why eventually, having a data scientist and a system built out to own your customer spend data will be a tremendous asset to the finance team.

Matt Heist added that your cohort data will also be valuable when presenting ARR up to your investors and board.

“Investors were saying to us ‘how are you going to convince us that you’re in good shape for the next few months or years?’ We kept going back to the cohorts, just continuing to look at usage history and segmenting it as best we could, usually by company size, and then by region.” – Matt Heist, Slack

Integrating ARR into Sales Compensation Plan Design

The amount of consumption-based revenue your business generates will almost certainly impact how you measure and compensate your sales team, said Karan Batra, who has led the sales compensation efforts at Fastly and Attentive. The key is to align the incentives of your sales organization with your consumption model goals. Typically, that involves activating new customers and driving ongoing consumption. 

“It’s often best to develop multi-layered compensation plans where you’re establishing a monthly run rate based on ACV and then continue compensating on usage above that run rate for 12 months. Depending on your business goals, you might also want to increase compensation based on the size of the customer or with different weightings for new versus existing customers.” – Karan Batra, Attentive

Ben noted that it’s important to install checks and balances on a quarterly or semi-annual basis for expansion-based comp models. He pointed out that bookings growth can become distorted if, for instance, customers are buying the same products repeatedly, but your system is counting them as new billings.

Similarly, pay attention to the on-ramp period with new customers. If usage tends to start slow and then increase after a few months on your product, Karan suggests measuring the highest 12 months of usage out of a 15 month period.

In Summary

ARR can be a useful business health metric for consumption-based businesses, even when measurement straddles between art and science. According to the finance leaders with experience measuring, reporting, and using ARR in sales comp models, the best approach centers around tracking customer usage and intensely-scrutinized cohort data. 

Interested in joining conversations like this one? Apply to join The Circle and gain access to a private leadership community where you can give and get insights and support from other growth-stage executives.