Server Branding 101: Big Name, Big Game?
June 2, 2008 Timothy Prickett Morgan
|
For kicks and giggles a few weeks ago, I spent some time monkeying around on the three top search engines seeing how many pages out there on the Internet relating to various IBM server brands and operating systems were being indexed with those terms. As I fully admit, this is a thoroughly unscientific way to try to gauge the relative importance of brands, but it seems to me to be as valid as doing a random poll by telephone–and has the virtue of being done in a few minutes. Now, let’s compare all the big server brands and their vendors. If you missed that story, which was called A Word Cloud of IBM Server Brand Names, the basic idea was simple enough. I trolled Google, Yahoo, and Live Search (from Microsoft) for old and new IBM brand names and counted up the number of pages that were indexed with these brands. Because these search engines come up with radically different page counts, I averaged the counts from all three and figured that’s representative. . . of something. I then created a bar chart and explained that this was very hard to read, and so I created something called a word cloud that better represents the relative importance of IBM’s server brands. With a word cloud, you use color and the size of the type to convey the relative importance of data in a set. What it showed, if you recall, is that the AS/400, iSeries, and System i brands were the strongest ones that IBM had in its portfolio, at least as far as Internet page counts were concerned, and that the 2000 rebranding (iSeries, pSeries, zSeries, and xSeries) stuck pretty well compared to the 2005 rebranding (System i, System p, System z, and System x). This week, I did two things. First, I did page counts for the server maker’s key brands: their very names or the abbreviation that they are most commonly identified with. As you might expect, IBM is a pretty good brand out there on the Web, but by virtue of the desktops and laptops that most of us sit at each day and the operating system that most of us use on those machines, Hewlett-Packard, Dell, and Microsoft are strong brands; Sun Microsystems, Red Hat, and Novell are not slouches in the data center, either, but are less well known on the Web. Here’s how the page counts of server vendor brand names (for players selling hardware as well as operating systems) stacked up:
It comes as no surprise that the Microsoft brand comes out on top, even though the company’s own Live Search engine appears to do an abysmal job indexing Microsoft’s own name. HP and Dell had almost as many pages indexed with their brands as Microsoft, and IBM was at about a third the level of Microsoft. Sun, Red Hat, and Novell had relatively small numbers of pages. Suffice it to say, on any word cloud rendition of these brands, Microsoft would dominate, followed by HP, Dell, a relatively small IBM, and the others would be dots. To collapse this data down a bit, as you can see in the table above, I then did a set of searches with the brands as well as the word “server” being required on the page. As you can see, this requirement for “server” to be on the page is the great equalizer. But even still, Microsoft would be the larger font, followed by HP, then IBM, then Dell, and then the others would be very tiny indeed. It is late Friday afternoon as I write this, and the traditional pizza is on the way as the kids are coming back from taekwondo practice, and so I skipped making a word cloud of these vendor names. Besides, I am more interested in actual server brand names. So let’s expand the table of IBM brands from two weeks ago to include the brands from other vendors. Here we go:
I stacked the server brands from HP, Sun, Dell, and Fujitsu-Siemens on top and added in the IBM BladeCenter blade server brand, which I forgot to include in the story two weeks ago. (Sorry about that.) I don’t know what the deal is, but Microsoft’s search engine thinks there are a lot more pages with “BladeCenter” on them then the others do, and I imagine there are a whole lot of places popping up selling knives or renting skates then I asked it to find. I forced it to find the term “BladeCenter,” but I have this feeling Microsoft’s search engine is not doing what I asked it to. Similarly, Live Search seems to be finding far too few “ProLiant” references on pages compared to Google and Yahoo. Which is why I am averaging numbers from all three. (I believe the statistical principle at work here is called it all comes out in the wash.) I like bar charts more than tables of numbers, so two weeks ago I created a bar chart for the IBM brands, which looked like this:
This week, here is a similar bar chart with the server brands I added to the mix:
The scales of these two charts are different, so bear that in mind as you look at them. The word cloud for the brands, shown below, really brings it on home in a lot less space than the two bar charts, and it also shows that the power of the brand X64 server brands far exceeds those of RISC and Itanium architectures–again, if page counts on the Internet are any indication.
Compaq, you will recall, got into the server racket in late 1989, a year after the AS/400 was launched and just as the RISC/Unix market was taking off, with the Systempro line. These machines featured Compaq’s own EISA bus, not the MicroChannel bus IBM tried to make a standard (and failed to do), dual processors, and eventually RAID disk controllers. The ProLiant brand was adopted by Compaq in 1993 as it moved into the data center ever so slowly, and in the following 10 years, driven by the adoption of rack-mounted servers in 1998 and the dot-com boom, Compaq had shipped 8 million ProLiant servers (including towers and racks) in the first decade of the brand. The cumulative shipments grew to 10 million by June 2005. It is probably a safe estimate that around 14 million ProLiants have been shipped to date, and that maybe half of them are still out there doing useful work. I think it will be a long, long time before HP changes the ProLiant, Integrity, or BladeSystem server brands. Dell seems to pick a name and stick with it, too. These are companies that are focused on sales, and they realize how disruptive name changes can be. IBM and Sun are changing server brands all the time, unless you admit that people still call two important machines at Big Blue “the AS/400” and “the mainframe.” Draw your own conclusions. As an IT journalist, I spend a lot of time with people in the marketing departments of the world. They are generally good, intelligent people. (I talk to a lot of CEOs, presidents, and founders, too.) But these marketeers can only be as good and as useful as their organizations allow them to be. I think that a long time ago, we had people who were involved in sales, and somewhere along the way, sales got split into one thing called “sales” and another thing called “marketing.” One involves a lot of responsibility for bringing in money, despite the shenanigans imposed by the other, best I can figure. It seems to me that people created marketing for two reasons: to push something that used to have a pull or to create the initial conditions that will set up the pull, and to give a more hoity-toity name to what is still, no matter how much you don’t want to admit it, sales. The trick, it seems to me, is to stop being ashamed about sales and do a whole lot less of the formal “marketing” nonsense, like constant rebranding, and to engineer servers and platforms that make your customers happy. That is the best way to get the point size up in that word cloud above. RELATED STORIES A Word Cloud of IBM Server Brand Names It’s Official: Now We’re Power Systems and i for Business IBM Changes Name Back to AS/400, Promises Return to Glory, TV Ads Progress Is Our Most Important Product IBM Aims for Server Expansion in 2008, Including System i Reincarnation A New Year, A New IBM Systems and Technology Group IBM Creates New Power, SMB Server Divisions
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
