Pre-packaged or Redundant? Reinventing the App.
As we reflect on AVIOS/SpeechTEK, the second major show of the year, the pre-packaged apps mantra that was sounded at TVUI continued to resonate on the floors of the exhibition floor in San Francisco. The echoes off the walls may have had an ironic ring for some, as they washed over thousands at the accompanying VSLive! event, where developers are being exposed to speech channels for their apps.
If it sounds like we’re not totally sold on the pre-packaged apps strategy, you’re right, we’re not. Opening challenge: there are thousands of pre-packaged apps for self-service that are out there today—all great candidates for speech-enablement. They’re known and vended by brand names like Siebel, PeopleSoft, SAP, Oracle, Great Plains, Best Software. And all those obvious verticals that are being targeted for speech apps, whether it’s utilities, financial services, government or telecom?
There’s a vertical solution from many of the ISVs just mentioned, complete with an ecosystem of VARs, SIs and user groups that know the back-end down cold. The first objection we hear is that ISVs don’t have the time of day for speech. While that may be true, you can look at the VAR and SI channel and ask yourself, aren’t they in a position to care – and act on the opportunity? The pushback to this objection extends beyond the specialist VAR/SI ecosystem – since many of these ISV deployments are hooked into significant CTI superstructure, we have to look at Genesys, at Aspect, at Cisco ICM and InIn and ask ourselves – do they get it when it comes to speech?
There is abundant evidence – much of it at AVIOS/SpeechTEK, that they do. So what really is going on here? On one hand, the industry is preaching open standards. On the other hand, it is acting like there are no third-party developers to take the basic building blocks and tooling and plug them into the existing IT infrastructure.
When the OEM strategy is eschewed for going direct to customers with turnkey solutions, there is clearly a regressive business model and a conceptual train wreck if that turnkey solution is:
(1) redundant with existing functionality within the enterprise; and
(2) predicated on open standards that invite participation by other stakeholders.
Having rehearsed many of the arguments with protagonists in the market already, we can predict the conversation that will ensue when you buttonhole us at the next event: "it’s all a question of semantics," we hear, "we’re really providing the VUI best practices and embedding them in software that will connect to the ISV back-end via VoiceXML."
So what is really going on is that lacunae in the middleware layer which, properly speaking from a three-tiered perspective should be filled in by adaptors and Web services utilities at the app server layer, pointing forward to the client (voice browser) and back to the application (ISV) are being supplied by speech specialists anxious to build market momentum. This is how capitalism works, seeing the opportunity and getting there first with a useful solution before the entrenched powers-that-be get it. But, call it semantics if you like, positioning these backfills as "prepackaged apps" seems to us like more of the same "21st century silo" thinking that we’ve warned about in the past. We can understand path to revenue is an imperative for new players that need to get past proof of concept and show a return to investors, but we question whether the same energy being put into selling the concept of pre-packaging might be better rewarded with actually plugging into the contemporary enterprise’s existing software assets via an ever-growing and more open set of third-party adaptors from major ISVs. If it was, we think these companies would find much of the work they are doing already is coded up for Web self-service. Isn’t that what Voice Web was all about?
Whether it’s verification, search, payment modules or voice-enhanced routing, these are all useful utilities, but they are just that: utilities. In the coming year, as the prepackaged apps story is overtaken by other events and the awakening of major IT players like the ISVs and Cisco and yes, Microsoft, the ecosystem will (hopefully) sort itself out between companies that "get it" – and those holding collateral about packaged apps that all look very similar.
What will the smart ones who "get it" be talking about? Modular enhancements that gracefully instantiate and work with middleware, integrated not through hordes of expensive professional services teams but via Eclipse-based tooling that embraces VoiceXML, SALT, Java and other componentry without constraints (economic or technical) on the developer. They will also get the fact that the developer is in many cases a VAR or an SI that knows way more about the customer and her vertical than a speech specialist will. In the meantime, it’s safe to predict we will get a lot of pushback on this somewhat contrarian perspective, but if there’s one thing speech has taught us, it’s patience. The incumbent ISV’s and their SI ecosystems who have exacted hundreds of millions of IT dollars from enterprises to develop Web-based self-service will get around to the phone, and their coattails should be ample enough for those who are agile and ready to ride.
Mark Plakias is a partner and senior consultant for The Zelos Group. He can be reached at 212.366.0895.