It is all very well expressing opinions on which technologies are hot and which are struggling, but what is happening in the real world? It is hard to get an accurate picture – surveys tend to have sampling biases of one kind or another, and vendors rarely release sales figures. I’ve never been happy with the TIOBE approach, counting mentions on the Internet; it is a measure of what is discussed, not what is used.
Another approach is to look at job vacancies. This is not ideal either; the number of vacancies might not be proportionate to the numbers in work, keyword searches are arbitrary and can include false positives and omit relevant ads that happen not to mention the keywords. Still, it is a real-world metric and worth inspecting along with the others. The following table shows figures as of today at indeed.com (for the US) and itjobswatch (for the UK), both of which make it easy to get stats.
Update – for the UK I’ve added both permanent and contract jobs from itjobswatch. I’ve also added C, C++, Python and F#, (which hardly registers). For C I searched Indeed.com for “C programming”.
Indeed.com (US) | itjobswatch (UK permanent) | itjobswatch (UK contract) | |
Java | 97,890 | 17,844 | 6,919 |
Flash | 52,616 | 2,288 | 723 |
C++ | 48,816 | 8,440 | 2470 |
C# | 46,708 | 18,345 | 5.674 |
Visual Basic | 35,412 | 3,332 | 1,061 |
C | 27,195 | 7,225 | 3,137 |
ASP.NET | 25,613 | 10,353 | 2,628 |
Python | 17,256 | 1,970 | 520 |
Ruby | 9,757 | 968 | 157 |
iPhone | 7,067 | 783 | 335 |
Silverlight | 5,026 | 2,162 | 524 |
Android | 4,755 | 585 | 164 |
WPF | 4,441 | 3,088 | 857 |
Adobe Flex | 2,920 | 1,143 | 579 |
Azure | 892 | 76 | 5 |
F# | 36 | 66 | 1 |
A few quick comments. First, don’t take the figures too seriously – it’s a quick snapshot of a couple of job sites and there could be all sorts of reasons why the figures are skewed.
Second, there are some surprising differences between the two sites in some cases, particularly for Flash – this may be because indeed.com covers design jobs but itjobswatch not really. The difference for Ruby surprises me, but it is a common word and may be over-stated at Indeed.com.
Third, I noticed that of 892 Azure jobs at Indeed.com, 442 of the vacancies are in Redmond.
Fourth, I struggled to search for Flex at Indeed.com. A search for Flex on its own pulls in plenty of jobs that have nothing to do with Adobe, while narrowing with a second word understates the figure.
The language stats probably mean more than the technology stats. There are plenty of ads that mention C# but don’t regard it as necessary to state “ASP.NET” or “WPF” – but that C# code must be running somewhere.
Conclusions? Well, Java is not dead. Silverlight is not unseating Flash, though it is on the map. iPhone and Android have come from nowhere to become significant platforms, especially in the USA. Beyond that I’m not sure, though I’ll aim to repeat the exercise in six months and see how it changes.
If you have better stats, let me know or comment below.
No better stats but I’m curious why you only used the stats for contract IT work from itjobswatch?
Andrew
Mainly because it is very easy to get the stats. Suggestions for another site?
Tim
Well the permanent jobs stats are available on the same site. They’re the default listing are they not?
Andrew
Fair point – I’ve added a further column.
It’s brought out more Ruby which is interesting, and improved Azure though from a small base.
Thanks
Tim
Hey Tim,
I also thought it interesting how close Silverlight and Flash are in the permanent jobs and that WPF is higher than both. This reflects real world experience I’ve had that WPF is in fairly high demand in the city (London).
Worth noting that Python ranks higher in both the UK and US than Ruby.
Another interesting metric of “popularity” is to look at numbers of books published on a topic. That gives a slightly different perspective than jobs. The jobs numbers are probably a good “real world” metric though. 🙂
Google search trends is another way of comparing, but even less meaningful… (I guess less meaningful ways of comparing technologies are not what you’re looking for – but I think it is an interesting topic.) 🙂
I’d be curious to know where PHP ranked amongst those. I’m based in NZ and PHP developers are in high demand… C#/.Net, not so much. It’s a small country with mostly tiny companies (90% of NZ companies are < 20 people, 80% < 10). Most use open source tools and very very few (probably only the 1000 corporate software devs as opposed to the 40,000 smaller privately held software companies) actually advertise developer positions.
It seems to me that the sort of data you're using above to draw conclusions will reflect what (mostly very risk averse) large corporations will be using for inhouse or product development. I would suggest that might not be very useful when looking for developer opportunities or at emerging trends/disruptive changes which are usually led by startups and small companies (remember Google, Apple, and heaven forbid, Microsoft?).
I know it's hard to measure (most interesting things are), but any other ways of measuring the *real* developer mind/marketshare out there? For example, Apache is the world's most widely used web server, and PHP is its most widely used module… that would, for example, support the idea that there could conceivably be far more real opportunities for PHP devs than .Net/ASP/C# people, but they may simply not be available through mainstream channels.
Have you looked at other metrics, like demand for online language info, api sites (e.g. W3C Schools), etc.? Any sample surveys of languages used by members of computing societies around the world, etc.?
From my perspective as founder of a 12 year old pure-play open source development shop, it seems that all the interesting startups in the NZ markiet – apart from a few heavily funded ones who are using MS stacks, and I suspect are heavily subsidised by Microsoft through partner programmes – are building their technologies on open source stacks (LAMP, Python, Ruby on Rails, Perl, Java, C/C++) for expedience, cost, and practical purposes.
@Dave
I’m not going to amend the table as other figures may have changed since yesterday, but for the record:
PHP
Indeed.com 20,483
itjobswatch.co.uk Permanent 4391
itjobswatch.co.uk Contract 950
Tim
Hi Tim,
Here’s my favorite query for isolating Flex developers from false “Flex” positives and “Flash” and comparing them to Java etc.
http://www.indeed.com/jobtrends?q=c%2B%2B,+java,+c%23,+python,+((flex+developer)+or+(macromedia+flex)+or+(adobe+flex)+or+(actionscript+flex)),+((silverlight)+or+(microsoft+silverlight)),+javafx,+objective-c&l=
Cheers,
-Brian
Also, Android passes iOS postings in 2010:
http://www.indeed.com/jobtrends?q=android,+((objective-c)+or+(iphone+developer)+or+(ios+developer))+&l=
Found this an interesting graph… especially when you take into account the corporate bias in the job listings…
http://www.indeed.com/jobtrends?q=%28%28php%29+or+%28drupal%29+or+%28joomla%29+or+%28symphony%29+or+%28magento%29%29%2C+%28%28jquery%29+or+%28javascript%29%29%2C+c%23%2C+python%2C+%28%28flex+developer%29+or+%28macromedia+flex%29+or+%28adobe+flex%29+or+%28actionscript+flex%29%29%2C+%28%28silverlight%29+or+%28microsoft+silverlight%29%29%2C+%28%28asp%29+or+%28aspx%29+or+%28asp.net%29%29&l=