(By Anthony Miller). My recent comment on SAP’s mid-market SaaS offering, Business ByDesign (see SAP caught in the SaaS margin trap) generated a lot of reader comment and a few press column-inches too (e.g. see Computerworld). My point that SaaS will be margin dilutive is not just a dig at SAP – I think it’s a truism for all ‘legacy’ on-premise software players.
Think about it – the clue is in the final ‘S’ – i.e. Service.
Simplistically (which is the way we analysts usually think!) SaaS has two parts; a software bit and a service bit. The software bit should be the same as any ‘on-premise’ software player, i.e. it requires R&D (lots to start with), maintenance/support, and a distribution channel. But to all intents and purposes, for an on-premise software player, that’s where their responsibility ends. They are not generally liable for end-user service levels other than (and I said this was simplistic) making sure the product meets functional specifications and that ‘performance’ (on a like for like functional basis) is at least no worse than the previous release. So, 30% operating margins are a good point to aim for; Oracle does 35%, SAP, 25%, Sage, 21%, and so on.
But what about the ‘Service’ bit?
A SaaS player is also responsible for end-user service levels. Now, how many IT service providers do you know with 30% margins? Actually, I can think of only one – Infosys (32%), a wild exception even among Indian players. Even payroll player ADP, arguably the ‘simplest’ case of SaaS (and surely much of this is, in effect, remote batch processing), gets 20%. Most IT services players struggle to reach double-digit margins; even Capita's are ‘only’around 12%.
I would therefore suggest that a 20% margin is about the best that a SaaS business can achieve (assuming ‘honest accounting’!). QED, SaaS must be margin dilutive for on-premise software vendors – even after huge start-up R&D and S&M costs have been written off.
SaaS is now a fact of life. Established software players (and investors!) need to get their heads out of the sand and recognise the inevitable effect on margins.
Friday 30 January 2009
SaaS margins – the clue is in the final ‘S’
Subscribe to:
Post Comments (Atom)
2 comments:
Anthony
I can’t disagree with your comments on margin reduction as a result of adopting a SaaS model – but it should only be margin reduction in the short- to mid- term. I would argue that the biggest effect of adopting SaaS is on cash generation (or lack of it), closely followed by investor expectation (of more rapid returns on R&D investments).
If a company can survive these for a 3- to 5- year period, by then the visibility of future revenue can allow a company to generate far higher margins than 20% - although many companies would use that to re-invest in more products and different markets.
Smaller, typically start-up, companies can build this into their business plans, ensuring sufficient cash and correct investor expectations are put in place in advance. The problem for established companies is to explain the financial impact of the move to SaaS in advance – and getting their investors to understand that it is a long-term move – that eventually should see an increase in margin (and a more predictable business).
Yes, there are (questionable) accounting steps that can be taken, and external finance options to help provide the cash, for the move to SaaS. But whilst these make the move to SaaS easier in the short-term, they tend to hit you 3 or 5 years down the line, and you don’t get the long-term benefit of the move to SaaS.
Cheers
Phil Benton
phil@systemsolveconsultancy.co.uk
There's a big difference between conventional IT services and SaaS. The former are typically highly one-off deals with little differentiation between vendors except price and reputation. They are also people intensive in sales, support and implementation. SME-targeted SaaS solutions are typically self-managing, one implementation/many instance, ie. capital intense but low operating cost. Add to that the opportunity for lucrative add-ons, and (once market share is built) margins should be high, possibly even utility-scale high for the best in class.
Jim Donovan, En Avant
Post a Comment