Marketshare Memes
It’s considered common knowledge that Microsoft Office controls 95% of the global office suite market, but this number comes from an obsolete market analysis framework and can no longer be considered accurate.
Building on this assumption without questioning it, the W.P. Carey School of Arizona State University writes “To Pay or Not To Pay: The World of Office Suites Opens Up.” The first paragraph begins,
“The ubiquitous Microsoft Office suite claims an impressive 95 percent market share. The software giant’s business division, which includes Office, brings in annual revenues of $16.4 billion. Yet since 2000, a free suite of software that includes spreadsheet and word processing programs similar to Excel and Word has evolved, with potential to chip away at Microsoft’s market dominance. Created by an army of volunteers, the open source OpenOffice.org has attracted a lot of attention but still claims a small percentage of the market.”
A few weeks ago, their sibling site at the Wharton School provided a few key words that clarify the whole situation.
Wharton’s analysis reveals the fallacy that 95% of users (PC users or office software users, pick one market) use Microsoft Office. Instead, the oft-quoted number simply measures that 95% of the revenue collected for the sale of office suites goes to MS (according to International Data Corp.). Thus, free products including OpenOffice.org and Google Docs are not measured by this statistic at all!
If you measure marketshare as the number of computers with the software installed, as a percentage of all computers, then Microsoft has far less than 95% and OOo has a good slice: Microsoft claims to have about 400 million MSO users, and some estimates place OOo users at 100 million. With these simple numbers, MS has 80% marketshare and OpenOffice has 20% (obviously excluding other players and overlap; these numbers could be refined).
When potential users see that MS has 80% marketshare and OOo has 20%, it will be clear to people that MS is vulnerable and competitors are growing strongly and quickly. This will accelerate adoption of OpenOffice.org, as those who are afraid to use it because they fear it’s not widely used, will see it is in fact very popular.
More accurate measurement tells a different story than what has been propagated so far, and the entire evolution of the marketplace could change because of it.
October 12th, 2007 at 10:26 am
This is what we’re talking about. Great article. It’s about time more people were aware of the FUD. We need constant re-analysis – I’m sure if we were to analyse the uptake of Operating Systems in the same way the analysis of the Games Consoles has been done we’d see lots more linux users. Lets compare Leopard, Vista, Solaris and Current Releases of Linux Distros and see where we’re at then. (As well has HPUX and AIX for the server market :p)
October 12th, 2007 at 10:45 am
[…] Horst has posted an entry, Marketshare Memes, over at his SolidOffice blog in which he explores some of the information regarding market […]
October 12th, 2007 at 10:56 am
[…] Market Share at 20%! Here’s a great post I picked up via the one of the OpenOffice.org mailing lists. It looks as though the 95% market […]
October 12th, 2007 at 11:27 am
I’ll put my OpenOffice.org hat on and show a bit of skepticism: those 400 millions claimed by MS are licensed users? How about the unlicensed ones?
I don’t know the source of the optimistic 100 millions OOo users, but does it count also StarOffice and Lotus users?
BTW, a few years ago some people wrote a “Strategic Marketing Plan” for OOo – http://marketing.openoffice.org/strategy/ (now it is forgotten) where the target for 2007 was… 20%
20% is considered by some a very important milestone, the tipping point where a player just can’t be ignored.
October 12th, 2007 at 11:44 am
The “100 million OpenOffice.org users” came to me most recently from Marc O’Brien, CEO of Projity, the company that created OpenProj (http://openproj.org/). I don’t see it right now in my quick perusal of the site, but I’ve got an email or two from Marc with that figure in it.
You can also look at the stats collected on the OOo website (http://stats.openoffice.org/), which indicates 98,300,000 + total downloads from that site alone.
These numbers do not include Lotus Symphony or Notes, nor do they include NeoOffice, StarOffice or other derivatives.
October 12th, 2007 at 2:10 pm
[…] MS Office – it’s not 95% as is often quoted – but may be closer to 80%. Read more to find out.read more | digg […]
October 13th, 2007 at 4:04 pm
Since my marketing project days with OpenOffice.org, I have seen a clear path to getting very good & reliable market-share data off of the BitTorrent trackers.
It is difficult to know the activity across the mirrors and gathering a reliable and accurate reading that way would be like herding cats from a unicycle…while it could be done, it represents a great effort accompanied by a likely high standard deviation.
It has been my understanding that through BitTorrent, usage data can be read & aggregated quite accurately and comprehensively from the trackers. From this information and its run-rate, a diligent person with under-graduate level statistics skills could extrapolate a snapshot of the size and composition of the OpenOffice.org user base — say all OOo 2.x.y.
100 million downloads = 100 million users is impossible. Consider how many times you yourself — as a member of the OOo project, for example — has downloaded and installed OOo on several of your own and other people’s machines over the life of the project and you might imagine the ratio of Total Downloads / Actual Users to be on the order of 1:5 or 1:10 or 1:35 — or that sort of magnitude.
In this light, 1:1 is what I would consider a carelessly optimistic assumption — if not disingenuous — based merely on common sense.
October 15th, 2007 at 1:07 am
Hiser’s argument for reducing the inferred total of users overlooks a simple point: Yes, some people will download OOo more than once, but equally lots will not download it at all, they’ll install it from a linus distribution. In fact I know heaps more people who installed that way than by downloading directly. The download figure is a good order of magnitude estimate, and other factors will both inflate and reduce the actual total. But you’ve got a reasonable number to start from. And, oh yes, the sales figures for M$ Office aren’t user numbers either: people do upgrade, you know.