SUP Dev Blog

In preparation for next week’s Google I/O conference, I wanted to glance through my thoughts from a couple months back, the day after the launch of the Google Apps Marketplace:

We have two camps emerging in the Platform-as-a-Service space: application-centric, and generic (application-agnostic). In the application-centric camp, we have Force.com, NetSuite, Intuit Partner Platform; in the generic camp, we have Amazon, Microsoft, Progress Software, RightScale, and a number of smaller process-centric players. (With the launch of Office Online, MSFT bleeds back a bit into the app-centric). IBM may join the generalist camp, and Oracle will join the application-centric as On-Demand Fusion Application adoption picks up.

With this Google Apps Marketplace offering centered around Google Docs, Google is positioning itself to integrate with Force.com, Intuit, and NetApp as the productivity core of these application-centric PaaS’s. In this context, it’s the ecosystem extension of the direct attack on Microsoft’s Office business, and a continuation of the underlying play against Microsoft’s OS platform. It foreshadows the increasing relevance of Cloud and service-based applications both in the consumer and enterprise spheres.

As to the degree of this relevance, my first take, after taking a few cuts at the forecasts, is that this platform is currently too small, and too loosely-connected, to strongly position itself as the core of a pure-Cloud, non-Microsoft stack—at least in the enterprise.

My first question is about the user base size. I’m frustrated by the Google marketing machine on this. The marketing numbers say 2 million companies and 25 million users. But going back to Henry Blodget’s breakdown in Jan 2009 [1], the forecasts were for between 200k and 800k paid subscribers by end 2009. In September, Reuters quoted a Google spokesperson at “in the hundreds of thousands”. (The Official Google Enterprise Blog this morning summarized it as ‘an increasing number’). Let’s say the recession drove sales to the upper half of the 200k-800k range: 650k subscribers in 2009. And growth through 2015 follows the Salesforce.com CAGR for 2007-2010 (33%).

Call it a $40 million business in 2010, and a $100 million business by 2013—for the productivity suite. So then what is the opportunity for the PaaS? To estimate the potential for an ISV on the platform, I would typically say that on a mature platform, a marquis partner might capture 25% of the user base for each of the application categories in the market. We use Concur at Sybase, and Concur happened to be one of the key partners at launch. So for Concur Breeze @ $8/user/month:

Assuming Concur as an organization grows at 15% CAGR (nice), capturing 25% of the Google Apps Marketplace expensing market would contribute 7% of revenue in 2010, and nearly 14% in 2015 (assuming no cannibalization or discounting). Not too bad. But none of the Application-centric Platform-as-a-Service offerings (Google Apps-inclusive) are anywhere close to 100% penetration of 3^rd party platform apps. Salesforce.com still looks to have the lead in this penetration metric with Force.com deployments on the rise, but I’m hard-pressed to assume anywhere near the majority of SFDC customers leverage 3^rd party Force.com apps. The Google Apps case is arguably a bit different, since the Productivity Suite is a more extensible platform in terms of use cases, but even extending that generosity, my ballpark estimate would be that <20% of the Google Apps customers purchase add-ons in 2010, growing to (probably less than) 80% by 2015. Then the forecast for Concur looks more like this:

It’s still very interesting to Concur, but it’s a long-shot to become a billion dollar business for Google at 20% revenue split. More telling about the direction of this new platform is its relevance to large partners.  For a partner as large as Intuit, the returns are much lower: at about $20/user, it’s barely a fraction of Intuit’s $3.1B annually:

***editor note:  I did a follow-up check on the status of the Intuit Goog Apps Marketplace offering Oct 28, 2010.  The pricing is a somewhat complex base cost of $39/mo, multi-state fee of $24/mo, and for the maximum 150 employees, approximately $1.25/employee/month.  The monthly total is approximately $250 for 150 employees–still equaling approximately $20/employee/annually.  You can access a pricing engine at http://www.iop4ga.com/gmp ***

Furthermore, Intuit made a very similar PaaS announcement with Microsoft at the end of January this year (“Intuit and Microsoft Join Forces to Deliver Web Applications to Millions of Small Businesses”). In that case it leveraged its own Intuit Partner Platform & Marketplace to deliver the Azure/Intuit SDK. The motivation for them to partner with Google is then far more geared toward the consumer Google Docs users—that’s how they get the potential user base up to 25 million—a large fraction eligible for TurboTax, QuickBooks, etc.

The question then is: will Google open the app store to ‘Standard’ edition users? Or will it make separate agreements to maintain extensions for QuickBooks, SFDC, (et al SaaS) directly into Google Docs? I think in their ideal world, the short answer is ‘both’. But we’ll have to look more closely at how Microsoft positions Office Online to determine whether specific integrations will be differentiating in the mid-term. Net Assessment:

• This set of Google infrastructure is not really generic cloud
• Google’s play in this case is to be the Productivity component of an emerging SaaS-based application stack (Salesforce, Netsuite, Workday, Intuit, Concur)
• This integration is not fully fleshed-out. There are still opportunities for native integrations of these potential partners’ applications, and it is unclear how or where the unification occurs. Today, integrations are happening on both sides. This is where Azure has a stronger message around MSFT-related SaaS and partners: there is a clear roadmap for integrating the SaaS stack. I don’t think Google has done a good job around Chrome OS in this regard.
• If they fail to be the unifying component of the stack, I think it is unlikely from an ARPU perspective that they will be able to drive greater than 4x in partner spend ($200 user/annually). As a result, forecast net platform ARPU at <$90. (20% split to Google)

Google could tilt the balance somewhat by:

• Deeply integrating Mobile into the Apps services
• Deeply integrating its mapping services as a component of Google Apps
• Rationalizing its social component services around Buzz and Wave

[1] http://www.businessinsider.com/2009/1/the-google-apps-revenue-myth-goog

Last 5 posts by Stan Stadelman

• MBO Best Practices - Video - December 12th, 2011
• The value of HTML5 in the Container - October 28th, 2011
• SAP developers describe first experience with SUP 2.1 OData SDK and NW Gateway - October 27th, 2011
• MBO Modeling: Best Practices - October 24th, 2011
• SUP 2.1 GA - October 24th, 2011