Anyone who knows me reasonably well knows that I tend to be a structured (albeit non-linear) thinker and that I have a penchant for semantics. Words have meaning and are not to be trifled with. Words with different meanings should not be used interchangeably; they have different meanings for a reason.
So when I got a call from a “adviser” to a growth stage business today promoting a “software company in the human capital management space” , I near about turned it down on the spot. You see “software” doesn’t fit in my investment thesis; “SaaS” does. You can imagine me saying:
I’m sorry, I don’t invest in software companies. I don’t like product company economics, including software license economics. I do however invest in SaaS. You see, I like recurring revenue businesses that have a long-term relationship with their customer. I like to sell something once and get paid many times over that long customer relationship. I like businesses with economies of scale, where the “next customer” has a higher contribution margin than all prior customers. I want to invest in businesses that have operating leverage and can be built into meaningful stand-alone businesses.
Thank goodness I’ve learned that all too many people use the terms software and SaaS interchangeably. So I asked the adviser.
How does the company deliver its solution, through software or SaaS?
His response:
SaaS.
Now we’re talking!
Old handles die hard
The adviser who called me isn’t the only person who’s made this mistake with me. Heck, one of my current investments, IP Commerce, once pitched me that they were a “software company”. I had to turn them around on that notion; “no, you are an on-demand service”. More recently, I had the a conversation with a cloud-services executive who made the same error; “no, you do not deliver a product, you deliver a service”.
I don’t begrudge anyone whose has used the words interchangeably. It is understandable in some ways; it is hard to let go of the past, and software is clearly the past when it comes to SaaS. People are clearly finding it difficult to shed the product-orientation of software for the services-orientation of SaaS. But let go we must! Why?
More than semantics
There are huge distinctions between software and SaaS. This isn’t just semantics. My Partner Doug Holladay would call this a ”distinction with a difference”. Which is to say that software and SaaS are two fundamentally different business models with fundamentally different:
-
underlying economics
-
key success factors
-
operating skill sets
-
capital formation needs
The point is that the differences between software and SaaS are more important than what they have in common. About the only thing that SaaS and software have to do with each other is that SaaS service-delivery infrastructure is software-based. But can you think of an on-demand services business; telecom, CRM, payments, cloud, securities trading, etc, etc. where that is not the case? I can’t (at least not one where I’d invest); its all software on the back-end. You don’t hear telecom companies proclaiming that they are “software based telecom”.
Yes, SaaS and software compete for market share in the vertical markets in which they overlap; which is just about every vertical market. And that is where the distinction really matters. Selling yourself as software when you deliver via SaaS is like pitching a netbook as a mainframe. It is a step backward. In most application verticals, SaaS has such overwhelming advantages against software from just about any perspective you can imagine (customer, operator, or investor), that any SaaS operator who utters the words “we’re a software” company is seriously degrading the value proposition of their solution.
So don’t do it. Be proud of the merits of on-demand operating model you’ve chosen. Sell against software; not side by side with it. You have a distinction with a difference!
OK, let me play devil’s advocate to stir some debate. Is a mainframe application a SAAS? It’s running on a server and delivered through en emulation terminal, emulator or host-web browser? Where a client emulator is used, it’s pulled down on demand, like Adobe is say!
What about Plug-ins like Adobe, Silverlight? Is this s/w or SAAS.
What about an app that will only run in Chrome (not IE or Firefox), using new Chrome specific features?
The only difference is these scenario’s is deployment. However, even most SAAS apps today deploys some tech down to the client (thin or fat) and there’s plenty of web 2.0 controls, applets, plug-ins that come over too.
Our “Software” company actually licenses to organizations using the SAAS model where it’s relevant (BPO’s for example). It’s pulled down as needed, on demand, as are the projects built with it. It’s built on .net
Rather that completely different, I think they are more the same, only some blurring of semantics.
Debate on, bring it on
I love a good debate. But, lets make sure we’re debating the same thing. The issues you raise have merit and are largely about how software is deployed. I’m prepared to acknowledge that the lines between software and SaaS can be blurry at this pure deployment level. And yes there are models around software + service with MSFT being the biggest proponent as we all know.
But lets isolate down to the service component of all of these deployent models. In the service model, most of the intelligence and operating complexity of your business is in your data center (or the cloud if you prefer); and as such, you have to manage a high-availability environment and likely get paid on a recurring revenue basis. You are definitively not a software company which packages its product, does periodic releases, gets paid on a license basis, enhanced with maintenance rev. Creating create software is necessary in both models. But in the SaaS model, it is necessary but insufficient for success; there is alot more to it.
Others have made the case that operating a SaaS business is more akin to operating a telecom comapny than it is to operating a software company. I think that argument has merit, despite the fact that software people reject it out of hand; heaven forbid they be put in the same category of folks at telecom people. But I throw it out there in the interest of stirring yet more controversy.
There is also an assumption that the “cloud” app you rent does exactly what you need. That’s rare to meet any enterprise’s requirements. Once you start down that “customization” path – assuming you even can – you are back where you started. Also note, many cloud apps have “plugins” that can be brought or “rented”.
I’m not sure it matters, if I rent a client app, that’s deployed over a WAN (VPN or not), that it’s not EXACTLY the same as a SAAS. Software + service – I think we concur. The development effort is the same but now, as an enterprise, if I won the s/w I effectively have more control.
I would also add, leasing software is not new. What is new, is the scale of the possible deployments for low cost. However, the same finances (Capital cost) debate is real. Enterprises do genuinely worry (as they should) about the viability if their SAAS vendor. The worry is exaggerated when the SAAS vendor stops “growing”!
[...] of the author and should not be attributed to Meritage Funds or any other person.Cross Posted At Non Linear VC. Posted Under : Analysis Tags saas themes Share this article: Permalink TrackBack [...]
Francis,
Regd. your mainframe argument, I will say if the service is delivered from the host and accessed by any software (read browser) on the client machine, it is SaaS. Though some aspects of mainframe might fall into SaaS category, there is a distinction between that and the current day avatar of SaaS.
Adobe and Silverlight plugins are all software and it is used to access service.
This is how I see SaaS. It is a service offered by some applications hosted in the cloud and which can be accessed using applications on local machines (PC or netbooks or smartphones or any device you use as client). These apps used to access the services can be browsers or RIA or, in some rare cases, a traditional application itself. The latter can only happen in a very rare instance where a big application in the enterprise workflow accesses some service like Salesforce. Eventually, I expect all of SaaS to be accessed using browsers as it is the only way to break open the restrictions added by different operating systems.
Hope it makes sense since I am in the “early morning” mode.
Krish, then we concur for the most part here. But, your point;
“Eventually, I expect all of SaaS to be accessed using browsers as it is the only way to break open the restrictions added by different operating systems”.
Does this make a browser(s), an OS? If everyone writes to the lowest common denominator (most do today and it’s IE) then your OK I guess.
I also have another take
I think this is the biggest game changer for the enterprise in tech in the last 10 years… More so than SAAS. Virtualization.
Virtual appliances give us the ability thus to run an O/S as effectively like any other container, does that make fat clients good again? Is the O/S now a SAAS
I think virtualization, in a very cost effective manner, enables applications to live longer, anywhere… Heck, you could argue, I could turn a fat client or DOS app, thanks to VM’s into a SAAS app
All great points Francis. Virtualization is clearly a game-changer as is its logical extension into the cloud.
But as for your point about software being available under lease, I’d argue that is a financing structure. A lease does not equate to a recurring revenue service delivery business.
Rather, what I’m focused on is the fundamental operating change that is required when a product (software) is now delivered as a service (SaaS). And that “operating difference” is where the “distinction with a difference” is most prevalent.