Why Competition Is So Bitter in SaaS: Oligopolies and Dominant Strategy Equilibriums'

Jason Lemkin

Perhaps the oddest thing about the Apple-Google “go thermonuclear” strategy to SaaS guys is that it is so odd at all.  Competition-to-the-almost-death seems the norm in SaaS.  Just look at Larry Ellison or Marc Benioff.  You can see the blood lust in their eyes, in every speech, in a way you never really saw/see in Steve Jobs or even Bill Gates, let alone Sergey Brin or Mark Zuckerberg.  As much as Bill Gates wanted 99% market share with Windows — I’m not convinced he really wanted to kill Apple and Steve Jobs.  Ellison and Benioff clearly want to kill their competitors.  And each other.  Every day.

Why is this?  I’ll admit at first I struggled to make sense of it, but once I came to Adobe and learned that the Omniture guys were just as competitive as Ellison, even with Josh James long gone, I realized it was The Way of the Universe.  But why?

I think it has to do with a combination of two basic factors:  Oligopolies in SaaS and the effectiveness of Dominant Strategy in SaaS.

First, why are SaaS companies “Oligopical”?  In other words, why are there generally just a few leaders that own almost all market share, with perhaps a few scrappy guys behind them?  I think we all know why SaaS is rarely monopolistic — there generally aren’t true market and network effects.  Just because I use Salesforce, you can still use NetSuite or Oracle.  I use Jive, you use Chatter, or Yammer, whatever.  There isn’t much data exchange.

But the costs to develop these apps becomes high, and it becomes higher as the markets emerge and in the beginning, often still remain small.  High costs + small market in beginning.  Look at the competition-to-the-death between Marketo and Eloqua.  Both wildly successful, one about to IPO, the other just IPO’d.  But even with all the Marketing Cloud hype, the total market even today for marketing automation software is probably less than $200m.  And it’s costing just the two of them almost that much to serve the market.  VC capital can distort this somewhat, by creating ‘room’ for an extra competitor or so — but that’s about it.  So you end up with room for only a couple of viable competitors, and some very small ones trying for scraps (Pardot, ActOn, etc.):

Chart from this PPT.

Second, the players generally are in no way interdependant – so pure dominant strategy becomes the norm.  As ugly as Apple-Samsung/Google is, the extremeness is unusual, because Samsung is also Apple’s largest vendor.  Most of the time on the hardware side, even when folks sue each other, it’s slightly gentler than it otherwise would be, because they’re often also each other’s customers.  Game theory says here, where there’s some level of interdependance, there’s generally an Invisible Hand that guides the competitors to an outcome that maximizes the greater good, or at least that involves a constant game of ‘chicken’, waiting to see what the other guy does first before expending effort.

But with SaaS, there’s no greater good and no interdependance among competitors.  Instead, pure dominance strategy works once the top competitors have efficient scale.  Dominance strategy states that your best strategy is irrespective of anything your opponent does.  So as long as there is capital to fuel it, SaaS competitors will try to enter every possible market niche, compete in every area possible of their market and for every single customer, spread as much FUD at they can, try to steal your customers even if the ROI is close to zero.  It doesn’t matter, because there’s no downside if the capital is there.  There’s no game of chicken, of wait and see.  Because there’s no ecosystem and no interdependance.

OK great, so what’s to do here?  I think the learnings perhaps are as follows, beyond the obvious of build the #1 best company you can (which needs no advice or discussion):

  • Understand if you are in an oligopical market.  If so, understand you probably cannot kill the #1 or #2 player once they have achieved efficient scale.  So try to kill them sub-efficient scale if you can.  But after that, recognize you cannot kill them.
  • If your competition has efficient scale and you don’t – run even faster.  Otherwise, they’ll be able to do everything and you’ll get cornered.
  • Learn to adopt dominant strategy tactics.  Once you’ve achieved efficient scale in your market, you need to be attacking every segment and every area aggressively, because your opponent will as well.  There’s no detente in oligopical SaaS.
  • Understand it never ends.  Because of dominant strategy, hyper-aggressive competition doesn’t mellow over time.  It simply increases, as more force and leverage can be applied.  If it feels like your competitor(s) are all around you, they’re in every deal now — get used to it.  It’s only going to increase as you get more successful.
  • Don’t let it get under your skin.  FUD, douchey partners, changing alliances, friends-turned-enemies … this stuff can be hard to take if it’s new for you (or even if it’s old hat).  Get over it, learn from it — and play a better version of the same game.
Published on October 2, 2012


  1. This is a terrific post. So if the SaaS market structure tends to be an oligopoly, what barriers to entry exist or can be built in SaaS to keep new entrants out?

    1. The barriers to entry usually are just (x) feature and platform richness and (y) brand / trust. Enterprise products become very complex and sophisticated over time. These don’t become true barriers at the bottom of the market where simplified feature sets work — but become serious ones at the top of the market as you scale. Salesforce has 100+ small competitors at the bottom of the market but has 95% revenue market share.

      1. Jason – I believe the relative merits / costs of building moats and toll paths are an issue here.

        A fast food outlet gives away disposable plates even though the store next door sells them. They undermine pricing or convenience barriers to adoption (in this case polystyrene plates, in Googles case say analytics or even free search).

        Similarly if you hold the gate to a walled pleasure garden (a toll keeper) you try to sell the ecosystem inside as being as great as possible (and FUDify the alternatives) but you charge for entry and low per ride.

        Alternatively you offer low entry (and exit) and charge the ride providers for their platform floorspace (market stalls).

        In the first case network effects dominate (Apple is a whole experience) in the second freedom of passage dominates (the whole world is a market) – This explains the difference between visitor revenue or device revenue. For my money the device is a commodity and walls can be brought down – I’m betting Google to hurt Apple rather than the contrary. If I was Apple I would want to bring outside pleasures inside my garden (office working maps etc maybe even wearable health) to do this I may have to open up a little (and then I am vulverable to free)

        So the platform is either a toll keeper (apple via hardware, MS software) or a conduit (commodity hardware and free platforms for traders right to market goods). Mix the models and the fun starts.

  2. Really helpful article – any additional insight you can provide for situations where you’re going in to disrupt a legacy market and the incumbent, single-tenant/ASP non-SaaS competitors start going after you by attacking the fact that you are multi-tenant SaaS? Also – what are your thoughts on going directly after the competition with direct comparisons and even naming names?

Leave a Reply

Your email address will not be published. Required fields are marked *

Share This