First, we look at the long-term trends from Indeed for our first 5, MongoDB, Cassandra, Redis, SimpleDB and HBase:
As you can see in this first group, demand has flattened for the past year to 18 months. MongoDB is still the clear leader. Cassandra stayed stable for a while, allowing it to build a lead over HBase which declined heavily during 2013 before flattening its trend. Redis followed the same general trend while maintaining its fourth position. SimpleDB has been in slow decline since 2012. It is possible that it will be removed in the next update.
Now, let’s look at Indeed for our second set of 5, CouchDB, Couchbase, Neo4j, MarkLogic and Riak:
First, a note on scale. In the previous graph, Redis is at 0.03%, while the top of this graph is at 0.01%. I did not include SimpleDB in this group mainly because of its age and declining trend. Given the scale of the graph, demand seems to jump all over. CouchDB seems to be leading but has dropped steeply from its peak in mid-2011. At just about the same current demand sits Riak, which also had a steep drop in the past several months. Riak does seem to be growing overall since early 2011. The major positive trend here is Couchbase, growing steadily since 2012. MarkLogic was in a bit of decline from 2010 until 2013, but has been growing since early 2013. Neo4j has been mostly flat since early 2012, probably due to its specialized nature.
Now, it’s time for the short term trends from SimplyHired for our first set of 5:
MongoDB seems to be increasing its lead in the short term, even with a dip in demand the past few months. Cassandra and HBase followed similar trends for most of the past year with Cassandra leading slightly. Redis continues a fairly stable trend for the past 18 months, lagging the leaders by a bit. SimpleDB shows little demand in a fairly flat trend.
For the second 5, the trends from SimplyHired are as follows:
The note about scale from the first set of graphs applies here as well. For most of the trends, there is a big bump in early 2014, except for MarkLogic. However, that bump was not sustained. CouchDB leads this pack, with a small gap over Riak. Riak shows the most unstable trend with several peaks and valleys. Couchbase is showing a solid growth trend, ending up at the same spot as MarkLogic. MarkLogic is also showing a good overall short-term trend, which could point to a positive direction in the future as well. Much like the long-term trends for Neo4j, SimplyHired is showing a mostly flat trend since the Autumn of 2013.
Lastly, we look at the relative growth from Indeed for the first 5:
This chart really shows the struggle of SimpleDB. While many solutions are growing rapidly, SimpleDB barely registers on the graph and looks to have a slight declining trend. Cassandra is leading the growth, just above 12,000%. HBase follows with near 10,500% growth. Redis trails a bit at just under 9000%. MongoDB, which leads in overall demand, is not showing the same type of growth, sitting just under 5000%.
Finally, we review the growth trends from Indeed for the second 5:
Unlike the general long-term trends, the scale on the growth chart is fairly similar. Couchbase is leading with a steadily increasing trend. MarkLogic comes next with a stabilizing trend, growing nice lately but really unstable between 2009 and 2013. Riak follows with a fairly stable, generally rising trend. CouchDB is definitely declining slowly falling to under 500% from its peak of 1000% in early 2012. Neo4j trails everyone, showing a flat growth trend for the past two years.
Currently, there are 4 major players and then a bunch of solutions fighting for attention. MongoDB, Cassandra, HBase and Redis all have solid demand and growth. This shows a very promising future for this segment of the industry. On the bad side of the trends we have SimpleDB, which looks like it is dying out. CouchDB seems to be floundering as well, even though its cousin Couchbase is growing nicely. The more interesting question is what happens with the other solutions, MarkLogic, Riak and Neo4j? The trends for those could be very telling for any solution trying to gain real acceptance.
After a slight delay we finally get to the third part of the programming language job trends. Today we review Erlang, Groovy, Scala, Lisp, and Clojure. If you do not see some of the more popular languages, take a look at Part 1 and Part 2. Lisp is included almost as a baseline, because it has had sustained usage for decades but never enough to get into the mainstream. Go and Haskell are still not included due to the noise in the data and the current lack of demand. Most likely, Go will be included in the next update assuming I can craft a good query for it. If there is an emerging language that you think should be included, please let me know in the comments.
To start, we look at the long term trends from Indeed.com:
Much like the previous two parts of this series, there is a definite downward trend for this group of languages. The trend is not nearly as negative as the previous two posts, but it is there. Groovy demand seems to be a bit cyclical, with new peaks every few months, and it still leads this pack. Scala has followed the same general trend as Groovy and keeping a large lead on the rest of the pack. Clojure has stayed fairly flat for the past two years and that allowed it to take a lead over Erlang. Erlang has slowly declined since its peak in early 2011, barely maintaining a lead over Lisp. Lisp is in a very slow decline over the past 4 years.
Unlike the previous two parts, the short-term trends from SimplyHired.com provide decent data:
As you can see, SimplyHired is showing similar downward trends to Indeed for Groovy and Scala. However, the Clojure, Erlang and Lisp trends look much flatter. Clojure has been leading Erlang and Lisp since the middle of 2013 and looks to be increasing its lead while the others decline. Here, Lisp is in a flatter trend which lets it overtake Erlang in the past few months. Erlang seems to be in a bit of a lull after a slight rise at the beginning of 2014.
Lastly, we look at the relative growth from Indeed.com:
Groovy maintains a very high growth trend, but it is definitely lessening in 2014. Scala is showing very strong growth at just over 10,000% and being somewhat flat overall since late 2011. Much like the other graphs, Clojure growth is outpacing Erlang, sitting above 5000%. Erlang growth is on a negative trend, falling below 500% for the first time since early 2011. Lisp is basically not registering on this graph as it is not really growing, staying barely positive. While most of these languages continue to grow, the trends for Erlang are not a good thing. Steadily decreasing growth for the past 3 years points to a language that will eventually become niche. Given that the overall demand was never that high, future prospects for Erlang are unpleasant.
Overall, these trends and the previous two installments make industry growth look flat. Granted, much of this is due to the breadth of languages being used, but even emerging languages are not seeing the same type of increasing growth. If you look at languages like Go and Haskell, the trends are not that much better. Go is definitely growing but Haskell is not. It is possible that both of these languages get included in our next update. Clojure growth is definitely interesting as it seems to be one of the few positive trends in all of the job trends. I would not be surprised if Clojure starts separating itself from the bottom pack before the next update.