Where Do Those IBM i Machines Work?
January 18, 2016 Timothy Prickett Morgan
We may live in a big data world, but a lot of the data is not worth the magnetic media it is stored on. This is particularly true when it comes to the IBM i installed base, which used to be more aggressively diced and sliced by the big IT market researchers of the world back it was called the AS/400 and when it had a $14 billion ecosystem across hardware, software, and services.
But we do what we can in a world without much data about our ecosystem, and that is one of the reasons why IT Jungle has lent a hand to HelpSystems, the software conglomerate based down the road from IBM‘s Rochester labs, as it tried to drum up respondents for its second annual survey of the IBM i installed base. HelpSystems is arguably the biggest independent software vendor in the market–it could be tied with or have surpassed Vision Solutions, the other big conglomerate, but we can’t be sure because both are privately held–but it cannot do this polling work by itself. HelpSystems, like the rest of the ISVs, has to find participants outside of its own installed base to give them intelligence about what is going on in their shops, which is why we and the remaining publications in the IBM i market helped out.
Last Wednesday I participated in two webinars hosted by Tom Huntington, executive vice president of technical solutions at HelpSystems, going over the survey, along with Alison Butterill, product offering manager for IBM i, and Ian Jarman, business unit executive for Power Systems Lab Services. Butterill and Jarman have been in the IBM i market for a long time, and with my 26 years and founding editorship of The Four Hundred I was basically the whippersnapper. During a discussion of the demographics of the IBM i installed base culled from the latest survey from HelpSystems, which you can get here, I gave myself a homework assignment to dig back through my archives and get some ancient data on the industry distribution for the AS/400 base back in the day. I found some data from 1995 that I had correlated to the breakdown of gross domestic product in the United States back then. Yes I know that is an apples-to-oranges comparison, since many AS/400 shops were outside of the United States back then and many IBM i shops are today. I just wanted to see if the customer distribution had any shape to the economy at large back then, and I repeated the experiment with the IBM i survey data collected by HelpSystems for the 2016 IBM i Marketplace Survey. In both cases, I would have much preferred to have compared AS/400 and IBM i site revenue to global GDP, but you go with the data you can get.
The first thing I realized is the way the data sources dice and slice by industry is not consistent, and you have to make some interpretations and consolidations to make comparisons, and these consolidate bits of data could mask underlying differences that would otherwise be apparent. This is why we have industry codes, after all, but governments are not standard on these and neither are surveys done by interested parties. Again, we do the best we can. So here is the data for 1995 versus 2015, when the latest HelpSystems survey ran:
If you look at the actual survey report from HelpSystems, manufacturing, distribution, and banking and finance continue to be the leading industries where IBM i platforms are kicking out useful work, but my comparative analysis of 2015 data with 1995 data that I have shows that the platform is much less dependent on some of these key industries than it has been in the past. The installed base is much more diverse and whole new categories have been added, categories that don’t fit into the ones that my data sources at Computer Intelligence (now part of market researcher Harte-Hanks) used. The interesting bit for me is that distribution–which includes both wholesale and retail operations in my original data–accounts for the same share of IBM i sites today as it did for OS/400 shops back in 1995. If you look at GDP figures, these distribution operations are a smaller part of the economy, and the IBM i platform has therefore increased its share of this market, by my simplistic comparison.
The reverse is true of manufacturing, which accounted for 35 percent of the base two decades ago, but which was about 21 percent last year, according to the HelpSystems survey. Financial, insurance, and real estate organizations are a much smaller part of the economy–this is hard to believe, but there is a plethora of services industries out there that dominate the American economy–and the IBM i platform has gained share here, too, even as manufacturing has declined. One more point on that: Manufacturing is a smaller part of the IBM i base than two decades ago, but it is also a much smaller part of the U.S. economy, too. Adjusting for a global economy, I think it would level out, since so much manufacturing is done in Asia and so much GDP is generated for that, but the IBM i platform probably does not have the kind of share of installations at manufacturers as it enjoyed 20 years ago.
I don’t claim this data is perfect, it is really just a way to talk about how the base compares to the economy as both are changing. As I said during the webinars, if the IBM i base looked exactly like the AS/400 base from years gone by, this would be the cause for concern. I also think that a lot of businesses that are in the Others category would have been called services two decades ago have ended up in that data. This is just a limitation of the data we have used.
Just for fun and just to give you a sense of what real data looks like when you see it, this is some internal marketing data I was able to wrangle out of Big Blue back in 1998. This data compares AS/400 sales for 1997, which totaled $4.2 billion at the time, to the potential market opportunity in 1997 as identified by IBM itself:
The numbers in the boxes show the share of the opportunity that the AS/400 platform took down. Back then, IBM barely dented the Others category, as you can see. We have no idea about revenues from Big Blue concerning the Power Systems platform, much less an industry breakdown or its effectiveness in taking a bigger bite out of the opportunity out there. We know less about how the IBM i platform as a piece of the Power Systems business is doing.
I think we are all a lot poorer because of the lack of information, and I wish we had a much better database from which we could all learn about our ecosystem. The irony is that in a big data age, when the tools are available to gather information for analysis and make correlations unimagined, companies have never been more stingy with their data. That’s because it is the new currency.