What Do You Care What Other People Think?

Earning our Place... Again

From my perpective, Microsoft established and maintained a market-leading position in the software industry with the Win32 platform by making far more money for others than it made for itself. 

This is only sensible - Intel did the same thing and are arguably, besides Cisco, the only other truly successful platform company (read Annabelle Gawer's Platform Leadership for a well-researched and written discussion of this).  Slywotzky's discussions of value networks are another way to look at the phenomenon - by adding layers of businesses that generate profit from your product, you earn a higher level of stability in your market.

We need to do the same thing again with software as a service - whatever set of strategies we develop need to focus on one thing at the core - make large amounts of money for the innovators who build new services & applications, for the channel that will make these available & relevant to the customers, and to the infrastructure players who will make it all actually work.  In a digital, low-friction marketplace of composable services the only way to succeed is by seizing the moral high ground - making a system that works for everyone else first.

Long Tail of Software: SaaS

Chris Anderson writes a fantastic post about the Long Tail in Software... not to rave excessively about software as a service, and for me the Long Tail is becoming something I see everywhere (both in SaaS and in Software Factories), but the tie-in is natural and well-written.  Go check it out, and catch Zoli Erdos' discussion too.

I think the specific questions this raises for the SaaS market go back to how to deliver an individualized service at global scale, with effective penetration of local opportunities.  How will a brand be applicable to very specialized versions or configurations of a software offering (i.e. who gets the opportunity to take Retail Management into rural China and how is that executed)?

As I've noted before, I don't find Multiforce or AppExchange credible coming from Salesforce.com (the problem is just too big for a player of their size/scope to tackle - time may well prove me wrong on this) but the core challenge that they're calling out is the right one, in my opinion - how do you deliver many different SaaS offers to a specific set of customers in a way that's relevant, leverages brand, and reduces cost of sale.  This software delivery segment needs a set of powerful marketing channels to make the business models work.

SaaS Franchises?

Will developers of SaaS applications enter a franchise model?

Given the high cost of scaling a SaaS offering compared to traditional software plays (requires much higher investment in operations staff, hardware, and software infrastructure) it makes sense for existing Service Providers (outsourcers like EDS, telecommunications/ISPs like British Telecom and Verizon) to own the operation of these services. 

For some set of SaaS brands, will it make sense to sell the software itself to service providers – a service provider licensing model?  These would then become well-known services added to those provided by a given customer’s existing SP (like call waiting or new cable channels).

How would companies establish and protect their brand experiences as this scales up?

SaaS and the Indirect Channel

Many SaaS providers will appear
There is likely to be a vast number of SaaS companies – we have already confirmed roughly 230 VC-backed companies and have in the range of 4,000 pre-VC startups who self-identify as SaaS companies.  Conservatively, we can estimate 25,000 companies at the height of this wave (using prior waves such as client/server as a basis).

SaaS Winners will create the Indirect Channel
Traditional software companies can tolerate their relatively high cost of sale due to the extraordinary margins which typify the software industry (80% gross margin).  SaaS companies have a higher cost base (operating & capital costs: hardware, infrastructure software, and operations staff) but demonstrate higher predictability of ongoing revenue.  As the initial disruptive effect on traditional software gives way to SaaS as the "new normal”, or the stability phase in value migration per Slywotzky, competition around cost of sale will become core to the segment.  The winners in the SaaS segment will be those who find ways to generate revenue for the traditional indirect channel (VARs, distributors, and SIs) and sell primarily through that model. 

I predict that there will be both branded and private-label offerings, and that these will line up with the channel model used by the vendor (brand will be important in driving pull for VARs and distributors; private-label will be preferred by SIs who are composing custom solutions for their customers).

Now, off to find some proof or disproof of this...

A Warm, Comfortable Feeling

I had an interesting conversation on the flight to Seattle today with a guy who runs West Coast sales for a large legal research company.  He lives in the Valley, has been in sales for 20+ years, and has been in successful startups in the past - very aware of current technology although his current business is law.  Really neat focus though - his firm focuses on ethics guidance for large companies (and yes, he was on his way to meet with Microsoft, one of his major customers).

He started using Salesforce about 5 years ago - to use his words, back when the application was really rough - but it was "so much better than Siebel" and "got better so fast" that he adopted it wholeheartedly.  I asked him what he did with Salesforce when he was on the plane or other disconnected situations.  "I just use their offline client - it works fine, no problems syncing or anything."  In response to my question about Outlook integration he said he just had Salesforce export the contacts to Outlook.

I told him about Smartcompany - how they use Outlook as a design center for their CRM smart client application connected to their SaaS offering, and are working on a future version that will deliver their functionality directly in Outlook - and his eyebrows shot up.  "Wow!  That would be amazing. I just tried out an application that looks like Outlook, and I got a warm, comfortable feeling because I felt like I knew how it worked.  This would be even better."  Conversation continued from there. 

The point here is that here is a lifelong salesman, user of CRM applications over the years who was an early convert to SaaS and Salesforce.com who would rather have that same service come through Outlook.

SaaS is maturing and the smart clients are coming.  Because user experience matters.

The Next Little Thing Isn't Free

Gary Rivlin wrote a great article in the New York Times today on the combination of Open Source and too much money chasing too few opportunities:

Any resemblance to 1999 is strictly isolated, more what might be called bubblets than outright bubbles. Open-source software, podcasting, social networking, security - these and other areas have been so hot at different points over the last couple of years that occasionally prices have inflated wildly.

Open-source technology companies in particular are in demand, with the widespread adoption of the software by corporate users. The group behind XenSource includes a pair of faculty members from Cambridge University and two software industry veterans. The young concern, in Palo Alto, Calif., received a first round of $6 million in January from Sevin Rosen and Accel Partners, also in Palo Alto, and Kleiner Perkins Caufield & Byers, on Sand Hill Road. But by midsummer, the start-up needed more money. "We wanted to scale this up faster than originally planned," said Mr. Sturiale of Sevin Rosen.

"I know of open-source deals being funded almost automatically right now," he said. Many venture capitalists, he added, are convinced that the open-source phenomenon represents a fundamental shift in the software market, "so they want to have a play there." That's what happens, Mr. Sturiale added, echoing others, when there are too many venture capitalists pursing a small pool of ideas.

The important thing to note here is that this is a leading indicator of the future of commercial Open Source, once thought to be the anathema of the Open Source revolution - VCs are putting money in only because they can see ways to get more money out, despite many failures in the recent past.  However, many people I've spoken with use the term "Open Source" very loosely and aren't always clear on why it's a good answer to a given problem.

I believe one of the reasons for this is the conflation of the term Open Source across multiple domains.  Don't get me wrong - I think Open Source makes a lot of sense, but there is confusion here.  Here's a stab at a useful set of distinctions:

1. Open Source as Development Model - leveraged software development approach using a small set of core committers with contributions from a broad community of developers and users.

2. Open Source as Commons - mutual investment into a designated shared property, where all are rewarded disproportionately to their involvement; originally described by Yochai Benkler and discussed extensively at FLORA.

3. Open Source as Business Model - disruptive low-friction distribution and marketing model which enables users to try software that was previously infeasible to acquire due to cost and complexity.

The industry has not yet adopted distinctions between these three meanings (and there are probably good cases to be made for additional meanings).  In the meantime, it behooves all of us as entrepreneurs, investors, and industry watchers to know what we mean by Open Source.

Many variations on the Open Source Business Model are currently in play and many more are coming - but the new wave of Open Source, as the VCs know well, is anything but non-commercial.

Venture Capital Open Source

Death of a Dialectic

Cliff Reeves mentions the strangeness of advocating for the power of the Open Source approach while working at Microsoft.  I agree that this is a nonsensical dialectic whose time has come and gone.  Even Ballmer was quoted recently saying "We don't compete with movements."  The common sense has been to take the Windows vs. Linux competition and turn that into an all-out "us vs. them" situation - both within Microsoft and the Open Source community.

Open Source is a perfectly valid - and frankly brilliant - model for developing software.  In an odd way, it's a combination of Friedmans' Flat Earth and Surowiecki's "Wisdom of Crowds", bringing together a highly intelligent, globally distributed self-organizing community that gets smarter the more challenges are thrown at it.  More on this in the future.

The point for this post is that people are still shouting but thousands of servers running Windows are hosting Apache, MySQL, PHP, Ruby, and hundreds of SourceForge applications - so it may be time to think about how OSS and MS can be less like oil and water and more like peanut butter and chocolate.

Microsoft Open Source

SaaS: What's Next?

There are many interesting discussions of SaaS these days, including this by John Loiacono of Sun - he focuses on a quick history of how we got here, and the business model core features.  Not dissimilar from most educated posts on the topic.

What I don't see anyone discussing is "what we can project about the future of SaaS?" given a reasonable interpretation that this is a new wave of application architectures - and that no matter how different the value proposition, it will share characteristics with prior waves.

Prior waves of application architecture (mainframe, classic client/server) had the following progression:

1) Application phase: delivered new functionality that solved a compelling business need, priced in a way that business owners who felt the pain could afford this solution.

2) Integration phase: with applications solving problems everywhere, organizations focused on efficiency at the data and security layers across these applications (eliminate multiple entry and multiple logins).

3) Knowledge phase: as integration problems became hammered out, managers and users became aware of larger-scale flows in their use of the applications - prompting richer client technologies and smarter integration that focused on process, context, and agility.

Note that these phases aren't discrete, but overlap in waves.  We're clearly in the early stages of the Application phase - there are hundreds of companies that are confirmed as true SaaS plays.  Compare this with 10's of 1000's of client/server vendors and I think you'll agree that this we are still in the early days with SaaS.

So what's next in SaaS?

Salesforce is the iconic application company, and started trying to solve the integration problem/opportunity by declaring themselves a universal platform ("Multiforce").  We'll see what happens there but this is a hard game to play and the road to platform is littered with the dead. 

We're already seeing companies like NewsGator and SmartCompany differentiating their offerings based on rich client integration (Outlook) and Grand Central saw the integration problem coming a few years ago (too early for the market).  Who is going to really get into the SaaS integration business and solve data access and identity integration problem?  Who is going build the platforms for better clients?

Because at the end of the day, we use browsers because we have to - not because we want to.

Microsoft SaaS