This caption coming from an archiving vendor, I would not fault the reader for thinking that this blog would again be about archiving and extolling the virtues of making production databases smaller. Yes, big (and getting bigger) production database size is not good, and one needs to archive little used historical data to make it smaller and make the business operations – not just the IT – more agile, more energy-efficient and more compliant with the need for immutable records management.

However, this will be about what I sometimes hear from customers that they feel a big vendor is always a safe bet. Having read the economist E.F. Schumacher’s 1973 seminal work, “Small is Beautiful,” I have confidently countered that; and happily, on most occasions, I have been able to convince them.

See the history and see what’s happening now. A venerable big institution like Bear Stearns went down. So did Enron. And now the two big mortgage companies are being bailed out by the US Government.

Let’s look at the Technology Industry. Once upon a time, long, long ago there used to be a very big company called DEC. Some of their products still survive, but not the company. During the best days of DEC, a small company – Oracle Corp – emerged to launch a RDBMS. Oracle later acquired DEC’s Rdb, and Rdb still survives on Oracle’s price list. Understand that in the early 1980s, Oracle – the small company then – had to fight the Goliath, DEC for RDBMS business on VAX/VMS.

And sometimes successful big companies enter a business only to get out of it later. GE, Honeywell and Xerox are good examples of this. All had at some point of time in their illustrious history been in the computer business. I daresay, only students of technology history remember this!

Having worked earlier for two organizations that were near start-ups when I had joined but became (and still are) the leaders in their field – and in parallel seeing what were then considered the Goliaths in the industry – either completely gone or struggling or out of that business, I would like to make the following observations:

• It is necessary for the small company to prosper and get bigger; there are no two questions about that. Too many small companies became history (and history does not even remember them) simply because they could not overcome the first minimum threshold of adequate customer acquisition and profitability.
• Once they acquire a certain size, it becomes necessary (but that’s not sufficient) to grow out of being just a niche player. See what happened to Netscape and Ingres.
• Perhaps the sufficient condition (unless there are serious management issues) is the fact that the big company just cannot afford to miss the technology bus. If they are late getting into that, it may be just too late to catch up. DEC missed Unix. Ultrix was just too late and not good enough.

Now coming to why GE, Honeywell and Xerox got of the computer business. It made good business sense. It was good business sense for IBM to divest its PC business. They exited in spite of the mainframe (GE/Honeywell) or Workstation (Xerox) or PC (IBM) markets being very good at that time because in this product line they were not number one or two in the industry; it was not major revenue or earnings earner; and they did not see this business as a prerequisite for success in their core areas. In fact it was the same reason why DEC gave up Rdb; the RDBMS market was very robust. But for DEC, this product line was not core to the success of their mini computer business.

So before deciding that bigger is safer, I would suggest looking at the two fundamentals of necessary (are they niche?) and sufficient (are they current on technology?) conditions of survivability of the big. And then one needs to ask if the product line is indeed strategic or even peripheral to their core business. Because, if it is not you can guarantee that sooner rather than later, it will be shed off.

It’s happening right now as we speak. In some cases, Small Is Beautiful. That’s why Schumacher’s book remains a classic even after thirty five years.