Another ERP vendor buys into middleware

The scale may be different, but the same market forces that fueled Oracle's acquisition of BEA have sealed another acquisition announced today: that of SOA middleware vendor Cape Clear Software by on-demand enterprise applications vendor Workday (also covered today here on ZDNet by Dan Farber and Dana Gardner). Increasingly, customers want their middleware bundled with the application stack. The less integration work they have to do themselves, the more they like it. Middleware is disappearing as a standalone software category.

Workday's acquisition sends many other messages too, but the rest of them offer far less comfort to Oracle [disclosure: Cape Clear and Oracle are both recent clients]. I recently visited Workday and learned that most (though far from all) of the company's account wins are PeopleSoft customers who have fallen behind with upgrades and can't stomach the cost and upheaval of moving to the latest version to get the functionality they need. They turn to Workday because implementation is rapid, the upfront cost is low and the on-demand model takes them off the upgrade treadmill for ever. (Read previous coverage of Workday).

Packaging integration into the proposition can only make the appeal even stronger. It reinforces that get-off-the-treadmill message, transferring responsibility from the customer to the vendor for implementing and maintaining integration as well as the application itself. In fact this always was the case — Workday has worked with Cape Clear as its integration partner since launch, so merging the two companies is merely a cementing of that working arrangement. But consummating the arrangement in an acquisition removes any ambiguity.

Just like a SaaS deployment, this private transaction will be complete in around a month's time. The contrast with Oracle's mega-bid for BEA couldn't be more stark. That deal won't close until much later this year, at which point two formerly bitter competitors will have to start working out how to consolidate two separate middleware stacks — a task that's likely to take another eighteen months to two years. It's just as drawn-out and painful as any Oracle implementation project. How many more customers will seep away to Workday in the meantime?

OK, Workday with less than 40 customers is small fry compared to Oracle's 280,000, but Workday has the greenfield momentum, carries no legacy baggage, and is already amassing an impressive roster of customers (its largest to date is food distributor Chiquita Brands, with 25,000 employees worldwide). In two years' time, I'd expect to see it become a much more substantial company, and a lot of people will be evaluating its ability to challenge Oracle and SAP. That, of course, is the bet that Cape Clear's CEO Annrai O'Toole and his backers have made with today's deal.

Where, by the way, does this leave other SaaS vendors and should they, too, be considering whether to package up integration capabilities with their on-demand offerings? (A topic I've previously discussed with O'Toole). Not necessarily by acquisition — Workday is something of an outlier in that it's one of the few SaaS vendors to focus on being a system-of-record vendor for the large enterprise (others such as Intacct and NetSuite have focused more on small and mid-sized businesses). For other vendors it may make more sense to use integration provided as a service within a platform ecosystem (as available within Salesforce.com AppExchange, for example, or from Saas hosting specialist OpSource [both clients, by the way]).

Perhaps there's also room for integration offered as a separate service, as recently announced by the likes of Boomi and open-source ESB vendor Mulesource. Or are those companies now themselves in play for acquisition? That's probably a topic worthy of a whole new blog post, but in the meantime, feel free to post your views in Talkback.