From Pieter Welten – Prime Ventures (@pieterwelten)

We all know that recurring revenue business models are (much) better than non-recurring business models for obvious reasons. Recurring revenues are special. They are important to potential acquirers, investors, (potential) customers, but also for budgeting purposes, risk / reward analysis, valuation purposes etc. Recurring revenue is predictable revenue that can be expected to be generated by a company in the foreseeable future. There are different types of recurring revenues and some are more valuable than others. Examples include multi-year contracts, (auto) renewal subscriptions, consumables models, knowledge memberships etc.

If it happens that you have recurring revenue streams you can extrapolate financial results into a future period (i.e. month, quarter, year) and get a relatively accurate idea of where the company will land in that specific time period. The number that represents the extrapolation of these financial results could also be referred to as revenue run-rate. Often, a company’s revenue run-rate refers to the expected recognised revenues of the next twelve months.

An example: Company X generates €100 in recognised revenues in January 2018 and operates an annual subscription business. Assuming none of its customers churn, you could argue that Company X will likely generate €1,200 in recognized revenues in the year 2018. Their revenue run-rate amounts to €1,200 in this scenario.

However, sometimes people talk about revenue run-rates when it does not make sense. They talk about revenue run-rates whereas their revenue streams don’t have a recurring nature at all! And unfortunately that doesn’t help anybody, not the entrepreneur, potential acquirer, investor or customer. Below I will give a couple of examples with regard to revenue run-rates in various sectors / different business models. The table below represents three monthly ‘P&Ls’ (Jan – Dec). Run-rate is function of monthly revenues * 12.

Enterprise Software Business
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec
Monthly Revenues 100 108 116 128 141 152 158 168 180 196 217 246
m-o-m growth %   8% 7% 11% 10% 8% 4% 6% 7% 9% 11% 13%
Run-rate 1.200 1.296 1.387 1.539 1.693 1.829 1.902 2.016 2.157 2.351 2.610 2.949
Consumer Internet Business
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec
Monthly Revenues 100 120 142 167 169 160 143 114 123 152 193 210
m-o-m growth % 20% 18% 18% 1% -5% -11% -20% 8% 23% 27% 9%
Run-rate 1.200 1.440 1.699 2.005 2.025 1.924 1.712 1.370 1.479 1.820 2.311 2.519
Hardware Business
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec
Monthly Revenues 100 160 70 70 118 353 176 194 58 12 477 239
m-o-m growth % 60% -56% 0% 67% 200% -50% 10% -70% -80% 4000% -50%
Run-rate 1.200 1.920 845 845 1.411 4.232 2.116 2.328 698 140 5.727 2.863

Enterprise Software Business

In my opinion it is absolutely fair to talk about revenue run-rates when you look at enterprise software businesses that sell annual or multi-year subscriptions to its customers. I am pretty sure you have read dozens of blogs about monthly recurring revenues (or ‘MRR’), churn, upsell and retention. Overall, when you look at cohort analysis you can objectively form a relatively accurate projection of the projected recognised revenues of the next 12 months.

A. Company has monthly recognised revenues of €116 and a revenues run-rate of €1,387. Assuming that all existing customers renew their existing contracts (i.e. no churn), you can assume the company will generate €1,387 in the next 12 months.

B. Company has monthly recognised revenues of €180 and a revenues run-rate of €2,157. Assuming that all existing customers renew their existing contract (i.e. no churn), you can assume the company will generate €2,157 in the next 12 months. In this scenario you can see that the company has grown its run-rate by 55% in two quarters and therefore you can argue that they will also will grow recognised revenues by 55%.

Consumer Internet Business

It gets a bit more complicated when we look at consumer internet businesses and discuss revenue run-rates. Lets assume two different consumer apps: A mobile app called ‘Foodly’ that allows you to order food and a marketplace called ‘Housely’ where you can list and sell your house. It is obvious that users of Foodly will use the app much more often that those of Housely; the average user probably orders 4x-10x per year food, but only sells on average once every seven year its house. As a result, you shouldn’t expect that a number of users that generate revenues in month X will also generate revenues in month X + 3 or X + 8. From that perspective, the revenues of Housely don’t have a ‘recurring nature’, but those of Foodly do. Since they lack repeat usage, for Housely it is paramount that they generate a positive ROI on marketing spend (or CAC) on its first transaction, whereas Foodly might require several orders before they start making a profit at user / diner level.

Hence, some consumer internet businesses can argue that they have revenues with a recurring nature and it makes sense for them to talk in ‘revenue run-rates’. You should keep in mind, though, that the revenue run-rates could still significantly differ over the months due to seasonality or increased marketing spend (amongst others).

C. In April, Foodly and Housely have each monthly recognised revenues of €167, but as we have seen it depends on the business (model) whether you can talk about a revenue run-rate of €2,005 or not. It is likely that none of Housely’s April customers return in the next twelve months and revenues will depend on new user acquisition. Their revenue run-rate has less value in my opinion. Foodly, on the other hand, might be able to demonstrate (with cohort analysis) that its current customer base will spend another €167 per month over the next 11 months and therefore you could argue that they will generate €2,005 in recognised revenues over the next 12 months.

D. Four months later, in August, Foodly and Housely both generate €114 in recognised revenues. With regard to Foodly, it could be the case that their customer base orders less food during the Summer months (hey, seasonality!) or that some customers have churned and moved to a competitor. The outcome is similar, Foodly’s revenues run-rate has decreased from €2,005 to €1,370, but as an investor I’d definitely worry in the latter scenario. In Housely’s scenario, their revenues (run-rate) could have decreased due to increased customer acquisition costs, less marketing spend, saturation of the market etc. At least it would not give me a lot of comfort that there is a high possibility that they will actually generate €2,005 in the period April to March next year as they might have said four months earlier.

Hardware Business

In my opinion, it is challenging to talk in terms of revenues run-rate when you operate a hardware business and don’t have a software component that generates recurring revenues. Lets assume a company called ‘Thermly’ that sells smart thermostats directly to its customers (i.e. B2C). If Thermly sells 100 smart thermostats in month X, they probably have no guarantee that they will sell another 1,100 thermostats in the following 11 months. There is a very high possibility that none of its existing customers will buy another smart thermostat in the years to come and therefore Thermly has to acquire new customers every month. Like Housely, it is paramount for Thermly to payback COGS and customer acquisition costs back on the very first transaction. I believe you get it when it comes down to hardware businesses, but check out E, F and G anyhow.

E. In March, Thermly generated €70 in recognised revenues. Their revenue run-rate would have been only €840.

F. In June, it happens that Thermly generated €353 in recognised revenues due to a bulk purchase of a large corporate that wanted to give all its employees a smart thermostat as a gift. In just one quarter, their revenue run-rate has jumped by 400% (!) to €4,232. Mind blowing isn’t it?

G. Another quarter later, though, in September Thermly generated €58 due to production and shipping issues. Their revenue run-rate dropped down to €698. As you can see, it is mission impossible to talk about revenue run-rates when you run a ‘hardware business’.

Concluding Remarks

So, my advice to founders, entrepreneurs, and managers who are raising external capital is as following: when you talk to investors, only talk in terms of revenue run-rates if it actually makes sense. In my opinion it definitely makes sense when you operate subscription businesses (monthly, annual, or multi-year) or when your revenues do actually have a recurring nature and you can demonstrate this with for instance cohort analysis. If that’s not the case you should erase terminology like ARR and revenue run-rate from your vocabulary. Why? You are either trying to mislead them or you don’t understand your business model. Whatever it is, it is not good and a seasoned investor will find out anyhow. I have experienced it myself. And trust me, investors will only value your business based on revenue run-rates if it is fair.