Sun claims “100 million copies of Open Office out there” – how many are virtual shelf ware?
I have just read…
…a rather startling claim ( from Simon Phipps of Sun that they have ‘shipped’ 70 million copies of OpenOffice.org with another 30 million copies coming from other sources – with the clear implication that 100m copies is in some way equivalent to a number of users approaching 100m. Taken at its simplest level, 100m users puts them at 20% of the office productivity suite business (as Microsoft claims 400 million users). Given that Microsoft got $11b in revenue from Office in 2005, does this mean that OpenOffice.org is taking away a whopping $1b+ of revenue from Microsoft- assuming that 50% of the OpenOffice users would have bought Microsoft Office otherwise.
Should Microsoft be quaking in its boots? While I have certainly heard good reports from some about Open Office, this seems to be a great example of how Open Source marketing can bend the normal rules of the software industry to the point where it is hard to figure out what the real story is. When Mr Phipps says that there are 100,000,000 copies of Open Office “out there”, he almost certainly means that 100,000,000 copies were downloaded or shipped on CDs. This is like a software company sticking an evaluation copy on the front of Dr Dobbs or BYTE and then claiming every reader as a customer. What matters of course is the number of actual sites using the product (and hence actually “investing” in it). It would therefore be reasonable to suggest that the actual market is a very small fraction of this and I would suspect that Sun has its own estimate of the actual market which it is choosing not to disclose for its own commercial reasons.
While it may be an easy shot to take at Sun and OpenOffice, the same problem pervades so much of the Open Source landscape. 10,000 downloads doesn’t mean anything if 9,990 were by students! To be serious for a moment, when evaluating Open Source for enterprise use you need to be very aware that the same rules do not apply as apply for vendor controlled software: For instance, reference sites may not carry the same weight as they are sometimes too close to the project and hence have their own intellectual pride and even careers are closely tied up in its success. In the worst case, what you thought was a reference site is closer to a sales pitch as the independent customer is in fact the primary author and sponsor of the project!
Before anybody takes from this that I am in any way anti-Open Source, I am not. However, any Open Source project from Open Office down does need to be evaluated using a different set of rules from traditional pay-for software.
Ronan
to be fair to Simon he never claimed there were 100m users, in fact he specifically stated Sun did not know how many people are using OO.o. I should have noted that in my post. You’re right though, enterprise customers do have to be aware that open source user metrics and usage scenarios are slightly different.
I was responding to a question from Matthew about why Sun is now offering support to OpenOffice.org users as well as selling StarOffice.
Indeed, I have no idea how many copies of OpenOffice.org and its derivatives are in daily use. I do not, however, that the it is extraordinarily widely distributed. As well as those downloaded direct from OpenOffice.org, the program is distributed with most copies of GNU/Linux, is being localised by 70 groups, has derivatives such as NeoOffice for the Mac, and so on. I’d guess (pretty randomly and for my own reference only) that 300 or 400 million copies have been distributed globally.
I’d also note that downloads are not the same as cover CDs – each download is a personal, non-trivial choice (it’s a big download), so the probability of use is high. Hence I am happy to guess that 100 million copies may be in use.
I explained that, no matter how you calculate it, the distribution of OpenOffice.org is wide enough to represent a significant opportunity which Sun, with its huge expertise in OpenOffice.org, would be foolish to ignore. I make no assertions in relation to other products’ markets.
Simon,
Thanks for taking the time to comment. I appreciate that it is not easy to calculate real usage levels for software so widely and freely distributed. However, I feel that this does put some onus on anybody making claims that end up in headlines to attempt to make a reasonable estimate of actual usage and in the worst case go for a conservative estimate.
Please note that I am not singling out Sun with this regard – it has become normal practise when discussing open source to focus on downloads over everything else. And I know that I must be counted as a user of many products that I have downloaded and never used.
To comment on your final point – which I totally agree with and was of course the real point you were making – Sun clearly has an excellent opportunity to capitalize on its expertise and investment in Open Office and it sounds like you are seizing that opportunity.
Ronan
Well, I consider 100 million conservative
And the headline is usually beyond the control of the writer (being selected by the editor in most publications), let alone the interviewee.
But thanks for the comments. My goal here is to ensure that Sun makes a modest return on its huge investment in OpenOffice.org so that it can be sustained for everyone’s benefit – even corporations can’t afford to make huge, multi-year gifts!