Given the headlines of accounting industry irregularities lately, no news is probably good news. But at 6:00 in the evening yesterday there was something late breaking: IBM’s acquisition of PwC. Yes, $3.5 billion for the same outfit that HP wanted to throw $18 billion at barely a couple years back.
What a difference a day, or in this case, a dot com implosion and an accounting industry shakeout makes. But more has changed than the price alone.
When HP pursued PwC, dot coms were still hot. At the time, we thought the idea of a computing vendor buying a consulting firm was risky, because the assets of the consulting firm — people — were only as firm as the provisions of their no-compete clauses. But today, with consulting firms laying off people, buyers like IBM don’t have to worry about retaining staff, they can probably cherry pick them. More importantly, the fit is much better than HP because IBM is already highly experienced at enterprise consulting. If anything, PwC gives them new vertical solution consulting depth.
Admittedly, bigger isn’t always better, as EDS found out the hard way with A.T. Kearney, an acquisition that had a different culture, target market, and sales cycle. IBM Global Services and PwC are a closer match, maybe in some ways overlapping. For now, bigger will likely be the rule at the high end — we expect this deal to drive more M&A among the EDSs, CSCs, and what’s left of the Big Five. Once we get through the current downturn (no crystal ball predictions here), less competition at the high end of consulting will create new opportunities at the price-sensitive lower end.
Way back when, somebody once remarked to us that the nice thing about standards was that there were so many to choose from. Not to knock standards — when it comes to the Internet, standards have certainly made the world go round. But when it comes to web services, we’re wondering if the world is going round and round and round.
At the risk of oversimplifying, the Internet worked because the standards were relatively basic and straightforward: TCP/IP for sending messages, HTTP for navigation, and HTML for presentation.
OK, web services are much more complex because they are supposed to provide a framework for automating business transactions. At minimum, that calls for standards specifying how to define a service, and how to design the envelope for sending it. And then, depending on your point of view, there’s the need for security, workflow, rollback, message delivery acknowledgements, directories and so on. Nope, this challenge won’t be answered by three simple standards.
And it won’t be answered by one or two organizations either. Beyond the Worldwide Web Consortium (W3C), the folks who originally gave us the web, just about every open systems group and vertical industry association wants a piece of the action.
A quick reality check is needed. Enterprises — the folks who are supposed to buy all this — are not exactly opening their wallets wide or clamoring for bleeding edge technology. And getting in the middle of a vendor food fight is the last thing they want.
At The Open Group’s recent web services symposium, held at a hotel on the far side of Boston Harbor, we saw a half dozen groups get up on stage and make nice with each other and spout repeated calls to keep the standards process open and democratic. But, as we were briefed about the various standards initiatives, we couldn’t help but stare across the harbor and think about the people who weren’t there, like Microsoft, which has authored many of the proposals now on the table.
Hopefully, the folks who call for openness won’t get too religious about the issue this time.
Last week’s Business Objects’ acquisition of Acta illustrates the bounds and limitations of packaged business intelligence (BI).
To recap, business intelligence complements core enterprise systems; while core systems report “how” a company is doing, BI systems can provide the “why” answers. Companies like Business Objects provide “front end” tools and applications that business analysts use to build their own query, reporting, and data mining systems. Back end tools like Acta are used by DBAs and developers to physically get the source data and transform it.
Outside the financial services industry, which largely competes on proprietary software, the rest of the business world clearly prefers packaged turnkey solutions. The latest example is BI, where ERP and CRM vendors alike have begun packaging analytic extensions to their core transaction systems, with BI tool and application vendors like Informatica, Business Objects, Cognos, and MicroStrategy packaging their own SAP/PeopleSoft/etc. add-ons.
The Business Objects/Acta deal is the latest example of this strategy. In buying Acta, which focused on ERP system data extraction, Business Objects is adding the back- end component so it, too, can offer end-to-end ERP analytics to rival SAP’s own Business Information Warehouse (SAP BW).
In this age of modest expectations, the $65 million cash acquisition was a bit of a bargain, representing but a modest chunk out of Business Objects’ $300 million cash.
Back to our main point, this deal shows the potential and limits of packaged BI. Yes, there is a sizable installed base of ERP packaged applications that provides a critical mass market. But most of the world continues to operate legacy systems for which there are no neat end-to-end solutions. Those cases will continue to require best-of- breed strategies that pair front end BI tools with either homegrown data extraction programs or tools from Acta’s former rivals: Informatica, whose strengths are with SQL database sources, or Ascential, whose edge is with legacy sources.
Consequently, there may be a modest rush towards packaged BI solutions. But for the rest of the world, BI will remain an area where assembly is still required.
A few months back, we mused about what really drives technology decisions. Assuming the product in question meets the specs, the criteria became deceptively simple: Does it offer investment and career protection? And when we commented on Java market developments last month, our conclusion was that the latest vendor moves were all about putting together critical mass technologies to provide the solutions in demand, such as portal or enterprise integration. They weren’t about Java.
So, the unannounced demise of WebGain, which became apparent at the end of June, reinforced each of these points.
To recap, WebGain was the Symantec spin-off that inherited Visual Café, one of the first Java tools. After the company was launched, it acquired several additional products covering design and modeling, database mapping, and some aspects of testing. The goal was to become a one-stop shop for tools covering most of the Java application development life cycle.
Although the company didn’t let on that it was going bust, the handwriting was on the wall for months. WebGain never integrated its tools and never delivered critical mass solutions, so there was no advantage in buying the full package unless you got it gratis. Furthermore, Visual Café — the centerpiece of its product set — used dated client-side technology with poor J2EE support (where the action is today). BEA, an early investor, walked away when it released its own WebLogic Workshop development environment last winter.
Obviously, listing WebGain won’t do wonders for anybody’s resume, but it won’t kill their career prospects or investments either. There are plenty of other Java IDEs out there, many of them open source. This is not like betting on the wrong appserver, database, ERP system, or enterprise server, because the skills are readily transferable.
Nobody is likely to get fired for buying WebGain.