Could, should Adobe follow in salesforce.com's platform as a service (PaaS) shoes?

IDG News Service's Juan Carlos Perez has published what is probably one of his best works ever, this one an interview of Adobe CEO Bruce Chizen (also, see Reuters' coverage of Chizen's webifications plans). The questions Perez asks are absolutely spot on. For example, he asks:

It seems that in recent months, the Adobe story has gotten more complicated than it was a few years ago as the company moves into new areas and technologies like hosted applications. Are you concerned that this might confuse and alienate existing customers?

Chizen agreed that there's more complexity than ever to Adobe's portofolio. Given Adobe's dalliance in hosted applications, Perez asked if online advertising would play a serious role in Adobe's future revenue mix (a great question given how that other shrink-wrap desktop software company has gone full bore with on-line ads).

One great factoid from interview: 70 percent of streamed video is played through Flash. Wow. Chizen brought this up when Perez hit him with the Microsoft Silverlight question. Chizen is clearly not one to let Adobe rest on its laurels. He's treating Microsoft as a serious threat but also reminds us that the Flash ecosystem has a 10 year head start on Silverlight. It will be a lot of ground for Microsoft to make up. Whereas the global footprint of Windows has afforded a certain distribution channel to Microsoft, much to the chagrin of the software giant's competitors, Adobe's global footprint is nearly the same as Microsoft's (Chizen claims 99.1 percent penetration). That doesn't mean that Microsoft can't even the score over the long run. But what it does mean is that Microsoft's natural distribution advantage isn't much of an advantage in this case.

Chizen also points out that Flash is on 300 million non-PC devices which left me wondering what on earth he was talking about. A little Googling turned up a story from March of this year about China Mobile running a handset download service on Adobe's Flash lite. China Mobile has 300 million subscribers. While it's not totally clear from the text of that story ---

Subscribers with supported handsets can access and download Flash Lite content from five different categories via their Monternet service which is available through their phone's web browser. With over 300 million subscribers for China Mobile there's a lot of opportunity for Flash Lite developers to sell their content to subscribers and we're really excited about the opportunity in China for growing a Flash Lite ecosystem.

--- I think it's safe to assume that not all of them have access to the Flash Lite-based service just yet. 300 million devices or not, the keys to Adobe's ongoing success lie in both the company's global cross-platform footprint and and the developer community that, for rich Internet apps (and even rich desktop apps) still sees Adobe as the go-to solution provider for the rich platform of choice (see When it comes to RIAs, is Adobe winning a one horse race?). So long as Flash is the go to video platform for giants like YouTube (it is) and big developers like eBay and Nike see it as the future to the interactivity of their sites (as they do), the Microsoft Silverlight team has at least a decade if not more of work cut out for it.

Speaking of decades, Perez very insightfully asks about Adobe's plans for software-as-a-service based architectures (where Adobe is dabbling now). Though he doesn't ask about it directly, you can't help but wonder if Chizen is also thinking about offering platforms (as opposed to just applications) via the SaaS model. That of course is one of the gospel's of salesforce.com.

A lot of people don't get the implications of salesforce's platform as a service (PaaS) strategy. But it's really quite simple. Today, a lot of companies run their in-house applications on J2EE or .NET servers and they have developers writing code to run on those servers. They're also running those servers themselves. Both platforms have baked in system functionality but not a whole lot of business functionality. In other words, J2EE and .NET are agnostic to the type business logic that runs on them. Meanwhile salesforce offers pretty much the same thing, only you don't need your own hardware (so you don't have to worry about the scalability of your apps), ID management and security are baked in (they don't have to be programmed in), and there's all sorts of hard core business stuff that's baked into the system for the developer's taking (in the Java and .Net worlds, you have to add that after).

In fact, speaking of Java and .Net, this sort of PaaS approach is absolutely ideal for runtime architectures which is, after all, what salesforce is running. A few years ago, when utility computing as a concept was starting to get some traction, I spoke with Sun CTO Greg Papadopoulos about running exactly such a hosted Java utility and wondered why, with so much of Sun's energy going into utility computing, Sun wasn't (a) building such a utility, and (b) in the case of Java, figuring out a way to commoditize billable units of Java compute power by allowing that utility to be equally powered by any server that could run Java (eg: Solaris, Windows, Linux, etc.). He liked the idea but Sun clearly wasn't anywhere close to having something like that. Surely, such a clusterable architecture could have profound implications for organizations wanting to run their own behind-the-firewall utilities as well as Sun who theoretically could have put itself into the multi-tenant Java hosting business the way salesforce.com is also now in the multi-tenant "progammable platform" hosting business. After all, who wants to worry about hardware and scalability?

Theoretically speaking, there's nothing preventing Adobe from moving to the same model where they run some giant multitenant Flash (or AIR) utility thereby saving companies all the trouble of setting up their own Flash laden datacenters. Like salesforce, such a utility would be a real attraction to the current crop of Java and .NET programmers if it offered some canned business logic the way salesforce.com does (not to mention that there's no reason that Flash developers couldn't simply call upon functionality from salesforce.com itself -- a point that salesforce.com veep of developer marketing Adam Gross made when I interviewed him visualforce.com). Or maybe there could be other forms of canned logic that Flash is well-suited to.

In the interview, Chizen says he doesn't see Adobe going the same direction Google is going with Google Apps. Adobe just acquired the Flash-based hosted word processor Buzzword but he sees that as being complimentary to Adobe's efforts to make it easy to collaborate around PDF files rather than being the part of some master plan to build a Google Apps-like productivity suite. But just supposing Adobe did offer Flash on more of a PaaS on-demand utility-like basis and just supposing that it did offer canned libraries of logic in Flash's areas of strength (eg: video) the way salesforce did with CRM and SFA, that would theoretically put Adobe in competition with the likes of YouTube or Brightcove. It's a move that Chizen will carefully have to consider since it might not be good for business to put yourself in competition with some of your biggest customers.

Chizen does wax a bit on the SaaS-based delivery of desktops apps but rightfully notes that quantum leaps in network bandwidth are not keeping pace with the quantum leaps in horsepower on desktops and notebooks. Said Chizen:

The desktop is a powerful, powerful machine in which to run applications. Broadband, as quick as it gets, is still going to have some limitations in the short term.

If the rest of industry was standing still, I guess Adobe could freeze its investment in client-side development and focus on offering the same functionality via a SaaS delivery model. But the industry isn't standing still and if Adobe doesn't continue to drink up whatever horsepower Intel has to offer on the desktop/notebook side, there are plenty of Adobe wannabe's who will.