Challenges to using revenue multiples

Steven Dunne

Steven Dunne
Senior Partner

Challenges to using revenue multiples

Ultimately all companies should be valued on their ability to generate future earnings and cash flows but multiples have commonly been used to short cut the process to assess company potential. Whilst earnings based multiples are clearly most closely related to the discounted future cash generation they are a proxy for, in the technology sector revenue multiples are more commonly used. Recent studies (e.g. by Catalyst and GCA) show that revenue growth is clearly the biggest factor in SaaS valuations.

Part of this popularity is practical as revenue multiples provide a basis for valuation irrespective of profitability and therefore can be used widely amongst venture capital backed businesses, but partly it is theoretically justified.  According to common business school teaching, revenue is less heavily influenced by subjective accounting decisions and is therefore a more reliable valuation basis than EBITDA or profit.

Within my accounting profession there would historically be acknowledgement that there is generally less room for manoeuvre on revenue than items lower down the income statement. Nevertheless, it has always been possible to have many different interpretations of the same facts and new business models such as SaaS have increased the variety of approaches. Revenue recognition is less consistent across companies than non-accountants might assume.

The new revenue recognition standard, IFRS 15, comes fully into force for reporting in 2018. As some large US tech companies have already adopted its US equivalent (ASC 606) or are flagging its impact, the potential for revenue to change has never been more obvious.  Uber are briefing investors on a potential halving of reported revenue whilst both Amazon and Microsoft will see the profile of sales change, higher in the short run but potentially more volatile as they are less able to spread revenue over time without clear contractual underpinning to that approach.

The Uber situation reflects the clarification of the Principal versus Agent judgement in the new rules. Previously Uber recognised the whole of the fare for rides as revenue whilst now they will be categorised as an agent in the transaction with the car driver the principal.  As a result, only Uber commission now counts as income.  Even with this clarification, different business models can cause difficulty in comparing across businesses or even between years. In our own portfolio, Scoota, a provider of a market leading adtech platform specialising in programmatic rich media, has successfully moved its business model from an old fashioned managed service to a much higher margin, self-serve proposition which provides greater customer retention and profitability but at the cost of revenue decline.  The managed service proposition saw them accounting for passed through costs as revenue but in the self-serve model only fees are revenue.  Revenue will appear lower in 2018 despite the business becoming a better, stronger, more profitable proposition.

SaaS businesses have taken many different approaches to revenue recognition but under the new rules the flow is apparently straight forward: (i) identify the contract and (ii) its performance obligations, (iii) determine the transaction price, (iv) allocate the price to the obligations and (v) recognise the revenue when the obligation is satisfied. In practice this is only straightforward if the contract is also clear-cut, whereas the reality of upfront fees, contract variations, rolled contracts, usage based pricing and other variations can create ambiguity. If SaaS businesses want to avoid a fairly painful audit process then this contractual side will, no doubt for many, have to be much better disciplined than at present. There is undoubtedly a cost involved in this but at Frog we are always encouraging our portfolio companies to invest in operational infrastructure to support scale-up requirements and this reinforces that need.

Our advice is that you introduce a clear sign off process for contract variations as well as initial contracts, managed by finance not sales, and they maintain the definitive list of contractual terms, updating as changes happen. Ideally you will progress towards having standardised terms and variation options so the categorisation of contract types is simplified; exceptions should require documented board level sign off.

It is tempting but dangerous to try to maximise apparent scale early through aggressive revenue recognition. At Frog, we encourage thinking ahead as to the nature of contracts and shape of the business and building the track record consistently. We advise that you don’t start recognising pass through costs at the outset unless that Principal position will clearly persist throughout your growth.

The only true reckoner for the potential of a business, particularly at venture capital scale-up phase where Frog operates, is how much have clients actually paid for the product or service, i.e. revenue collected and retained in cash.  Only cash is absolutely immune to accounting interpretation but even here there are issues to watch out for.  Remember that when companies get paid in advance they show the difference between revenue recognised and cash received as deferred income, and there is good reason why this is a liability on the balance sheet. The revenue has not been earned yet and the costs of delivery have not been incurred so when looking at cash generated and surplus cash for valuation purposes, deferred revenue should always be taken into account.

Hopefully the new rules will bring improved consistency but it will remain a fallacy that all revenue is comparable.  The logical evolution might be to utilise gross profit as a better basis for valuation than revenue but, especially in the software sector, the variation in accounting for cost of sales is great.

There are no easy answers, and no alternative to properly understanding the business model of any company being valued and those believed to be comparable. There is also no definitively reliable valuation methodology and multiple approaches should always be used. For those producing revenue numbers, consistency and transparency are the key. Always look ahead both in terms of policy changes and the nature of the business model to avoid the need for sudden changes of approach. Surprising changes in revenue recognition are likely in themselves to create uncertainty and potentially a valuation issue.

Steven Dunne


Steven Dunne