Open vs. Transparent: Did FOSS, Linux, and Open Source Get it Wrong?

First, there is still little interest in Vista in business; second, there is a credible effort to use Linux on all mainstream corporate desktops but not instead of Windows, in addition to it; third, Windows Server 2008 is being deployed widely in beta; and fourth, Linux interest on the server is declining.

Now, suddenly I think I’ve stepped into another dimension where up is down and down is up. I can’t yet talk about the Linux desktop effort in depth as it is still relatively secret, but I can say that it has more to do with securing and managing the desktop then it does with displacing Windows, and that the reason behind the choice had more to do with cost than almost anything else.

For the Vista stuff the lack of demand is being attributed to two things: Windows XP is seen as good enough, and Microsoft isn’t Apple and can’t do demand generation marketing to save its life. Windows 2008 server, on the other hand, evidently addresses a number of shortcomings in Server 2003 and, apparently, the benefits are actually exceeding the risk of deploying a beta for a large number of folks (that doesn’t happen often).

But, this time, I’d like to focus on the last, the comment that interest in Linux was declining, and explore for a moment why I think that may be the case.

Birthing Open

Ever since I first heard the term “Open” I’ve wondered whether a lot of otherwise intelligent people missed the message that had been consistently sent by IT buyers. That message was that they really didn’t want to be in the software business.

Now the problem that Open was supposed to address when it started to surge in the ‘90s was one, that to me anyway, seemed to be largely based on the lack of trust with a number of vendors, and while Microsoft was clearly one of the problem vendors they were far from the only vendor painted by this broad brush. At the time, Sun, IBM, Oracle and a broad cross section of very powerful software vendors were consistently showcasing expertise they didn’t have, promising features that didn’t exist (and often never would), and generally setting expectations for very expensive products that they had no real intention of ever actually meeting.

We were reporting that CIOs had the highest turnover of any top-level executive (thanks to the dotcom years they were, for a time passed by CFOs but that hardly made them feel more secure). And much of this turnover likely could have been tied to what was a cavalier attitude on the vendor side, with respect to both protecting their customers and ensuring their success.

There wasn’t that much concern surrounding software patents and intellectual property, except between vendors. And even between vendors this typically only happened when a patent troll started wandering around the industry, or a large vendor decided to mine their patent portfolio and go after competitors.

Most of the companies had trouble collaborating within their various divisions and the idea of collaborating broadly, while it clearly existed in parts of IT – and we spoke about it as a goal a great deal – practically speaking it wasn’t a driver. And, with the exception of the insane CIO who from time to time became convinced he could market his services to other companies, particularly from operations there was no desire to actually go into the software business.

Finally, hardware costs were dropping like a rock and software costs were either holding steady or increasing. As we moved in to this decade and a heavy focus on cost reduction hit IT, having a cost category that either didn’t move in the right direction or moved in the wrong direction was a big problem. Microsoft made this worse by picking this inopportune time to “fix” their licensing model, which resulted in a massive number of companies getting audited and finding out they weren’t actually paying Microsoft what they had agreed to pay.

I would argue that the problems that most CIOs were facing had nothing to do with whether the code was “Open” or not, it had to do with trust, and IT executives simply didn’t trust their vendors – and Microsoft was the poster child for this problem.

Open Source presented itself as an alternative and Microsoft became the target for much of the marketing and rhetoric that ran behind it. But, I think, the market is beginning to realize they’ve been had and that, in a number of ways things have actually gotten worse.

Open vs. Transparent and Customer Focused

I think what the market wanted was for the vendors that served them to be candid about their offerings. In short, to not over promise, and stop doing things to them and start living up to what they promised and do things for them. What they largely got with “Open” was the right to do it themselves, yet they didn’t even want, with some already existing exceptions, to do it themselves.

What IT folks now appear to be realizing is that Open doesn’t mean transparent and that the folks like FOSS who promote Open have their own agendas which, while clearly different, isn’t any more customer focused than a proprietary vendor.

In effect, named vendors who are now “Open” are generally not any more transparent then they were when they weren’t open, and they are often very selective with what is actually “Open.” In the end, I think, IT buyers are now realizing that all they did is exchange one set of problems for a new set and that Linux, in particular, is kind of just like UNIX but with less support, less software cost, but more labor costs associated with it. And they now have to worry more about intellectual property and license proliferation than ever before. In other words, complexity actually got worse.

By the way, I should be clear, I’m not saying that Linux is going into decline, only that it appears the growth rate is dropping off. But I think this means that buyers are becoming more realistic in their expectations.

Wrapping Up and A Little Advice

The best way to deal with a trust problem is simply to choose not to do business with companies you can’t trust. And for everyone else, make sure trust is backed up with solid research into whether the company can actually execute. It amazes me how many people trust companies to provide solutions that the providing companies own IT organizations won’t even deploy, and how few companies actually, for multi-million dollar deals, do any due diligence with companies who have been held up as reference customers.

If you are trying to deal with a trust problem, Open probably won’t be much help. You want transparency and to get that you need a relationship with someone in the company who has a clue. This is as true of RedHat as it is of Microsoft, and let’s be clear, sales and marketing departments are often not included in the “has a clue” group. You need a line manager – and one that is high enough so that they’re actually in the decision loop surrounding the critical product or service you are depending on to work.

Using Microsoft as an example, and I used to survey for this extensively, there are a lot of CIOs that trust the company less than any other firm; there is a near equal number that trusts them more. The difference is the second group has relationships with the company that solves the transparency problem (by the way, not all executive relationships work, I’m only saying that, for the companies who trust Microsoft, those relationships are working).

I’ll close with this thought: if you can’t fix a trust problem, it’s time to pick another executive relationship (it may surprise you to find out that some executives are empty suits, or worse, intentionally dishonest). This is particularly true if other firms aren’t having the same problem with your problem vendor. Or, at last resort, you need to pick another vendor and consider creating better criteria for selecting vendors in the first place.