I suspect the proof in the pudding will be who actually replies. I suspect I am bang on with who will and who will not. :D
I'd be surprised if they don't at least part reply mate. I suspect everyones waiting for someone to set the standard - go on @PhilNeal, once again set the benchmark.
I tried something like this two BETTs ago but not all replied and then we couldn't publish it and it got all complicated. Progresso failed, didn't get anywhere with RM , ended up being more hassle than it was worth.
Credit to Capita they were first to help out. Scholarpack and iSams also very friendly.
He's got no choice but to finish it and to a high quality - he knows what happens to people who don't complete stuff for me!!
There have been some discussions and developments with some suppliers, I will update soon about some extensions to the challenge that will make this not only more worthwhile for all involved but give a more rounded approach to the results. More details soon on this....
The website to host the results of the challenge is now live, although needs some work still!
However, the results from the past 4 years' Census and subsequent market analysis and calculations are now complete and live and available for view (or download if you want the full file). This originated on work done by @matt40k and I'd like to thank Matt for his input both before and during this.
Note that the figures are taken directly from the school census', I have not altered anything. If you find something is incorrect in the details please do let me know and I will alter as best I can (ultimately, blame the DfE!). I'm aware things like type/phase of school may be inaccurate in some records, I've done my best but will apologise in advance anyway (basically if you spot your school and there is a problem I will change it! I want this to be as accurate as possible) Also, the figures should be taken as a guide only not the definitive accurate exact value as I am aware not all schools have been supplied in the census for various reasons, as well there are new ones popping up all over so make analysing against past years somewhat interesting (plus, I have used calculations to match year on year data, I was not going to go through 84000 records one by one!). Again, school or supplier if you have a query I am happy to address. Enjoy!
Website is at: Market Statistics | Eduware Consulting
Coming soon will be the MIS Challenge survey pack, that will be freely available to see and/or use by anyone (ie as a base for your next procurement... I can dream :)).
Hmmm not sure I like some of the +/- 2012 stats. It feels wrong seeing new entrants entering as 500% etc, if it is the first time they are included in the count then they need to be marked as new (eg, 0 in 2012 and 5 in 2013 should be New Entrant, not 500% as it is misleading).
It's also hard to read. If you plan on showing year on year growth then you want it in chronological order, starting with the earliest and working up to the latest. Show the stats for 2012, then the stats for 2013 then the difference. :)
matt40k showed the same things...
Have you looked at the full Excel file which shows the year on year stuff. This page you see is a summary only of this year.
I think you're hitting the same issues I'm hitting, how to best present the data. The growth fact is correct and although it appears to be ridiculous it isn't worth completely overlooking - the point of it is to show growth or decline.
It's growth but it's skewed growth. Zero is a very hard number to work with, put it on a graph and see the mess it makes. :)
Take this set of stats: Browser Statistics It's easy to see that Chrome entered the market in Sep 2008 because it is the first entry, everything else is blank. They could have left it on 0% all the way to the bottom but that would be harder to read. Blank means blank, nothing here, nothing to see. 0% implies there but failing to make even 0.01%.
Likewise, if a browser had 100,000 users during the first period of stats would you really say it had 100,000% growth? I'd look at it and say "No, that's not right and something is wrong." Add it to a graph where growth for others is small, about 1-2% for some, and you would get a massive bar with everyone else a line along the bottom axis making the graph fairly useless. It also gives the impression that the new guy is dominating, when they are actually a handful of users. The following year would also look bad. Imagine the number of Chrome users went from 100,000 to 200,000. That's a 100% increase which is fantastic, but the growth would drop from 100,000% the previous year to a fraction of that figure at 100% making it look like they dominated the first year then plummeted the next making for a freefall in growth which isn't true.
Oh believe me I know 0 is a bugger of a number to work with.
Don't get me wrong I completely agree is looks odd, however to keep everything fair I couldn't put any negative weighting on that. Take a look at the downloadable excel file which will give some clarity, if you have an suggestions on other ways to present this I would like to hear. I did have a % market share change too after that but everything was so small it was pointless having it on there (it is in the file still).
Maybe you'd be better off showing userbase movement as raw numbers instead of a percentage? That way you could see X gained 11 users, Y had better growth at 23 users and Z showed a slight decline at -8 users. And even if W shows a massive haemorrhaging of -100 users it would still be relevant and visible!