The latest expert opinions, articles, and guides for the Java professional.

IDEs vs. Build Tools: How Eclipse, IntelliJ IDEA & NetBeans users work with Maven, Ant, SBT & Gradle


In our recent Java Tools & Technologies Landscape for 2014, over 2000 developers told us a bit about what JVM technologies they use and which tools are interesting to them. From that, we produced a gorgeous, professionally-designed, 60-page report (which you can download) on 14 technology segments. We recently dug deeper into stats about users of Spring MVC, JSF, Vaadin and GWT and the post ended up being really popular.

So we wanted to do the same with IDEs, but realized that we had so much awesome data that we needed to continue to break it down into separate tool segments from there!

So why does it make sense to segment by IDE?

  1. 97% of developers surveyed use an IDE
  2. Most developers use only 1 IDE exclusively.
  3. Most developers use the same IDE for a long time.

These factors lead us to believe that you can potentially learn something about developers based on the IDE they use (or choose). What we’ve done is extracted data for each IDE users and compared it to the average using Microsoft Excel pivot tables. What we are seeing is how a specific IDE users compare to the entire survey population, and seeing if anything significant can be seen. Let’s get into it, starting with Build Tools…


IDEs vs. Build Tools

Given the deep integration and love-hate relationship between users of different IDEs and build tools, we thought about starting here and covering Eclipse, IntelliJ IDEA and NetBeans, which comfortably take up over 90% of the IDE market. Other tools, like Spring Tools Suite, MyEclipse, IBM RAD, JBoss Dev Studio, vi/vim, emacs, etc didn’t make up significant portions of the overall community, so to make things simpler we didn’t cover them at this time.

Eclipse (48% IDE market share)


As the most commonly-used IDE at 48% (when increases to 55% when including distributions like Spring Tool Suite, MyEclipse and JBoss Dev Studio), we expect Eclipse statistics to more or less mirror the average. However, we can see a few differences when compared to the entire survey population. Most noticeable are:

  1. Ant (+/- Ivy) usage (22.8%) by Eclipse users is significantly higher than the average (16.5%), giving them highest % of Ant users in the industry. In light with Ant’s visible decline in use that we’ve seen over the last years, this indicates that Eclipse projects are frequently mature.
  2. Maven usage is very aligned with the average—also, the % of respondents using No Build Tool is also quite close to the average.
  3. Usage of Gradle (5.8%) and SBT (0.8%) by Eclipse users is considerably lower than the average, which is 11% and 2.5% respectively. This may indicate that Groovy and Scala users are perhaps choosing another IDE…

IntelliJ IDEA (33% IDE market share)


Our survey distinguished between IntelliJ IDEA Community Edition (free) and IntelliJ IDEA Ultimate Edition ($$$), and for the sake of the graphics we combined the weighted scores for both versions. IntelliJ IDEA is the fastest growing IDE in the market, and according to 49% of respondents, the “most interesting” IDE available. Here is what we see here:

  1. Gradle usage (20.4%) by IntelliJ IDEA users overall is nearly double the average (11%), which shows a clear preference for newer technologies and dabbling with alternative JVM languages, like Groovy.
  2. SBT usage (5.2%) by IntelliJ IDEA users overall is over double the average, and it skyrockets to a whopping 11.6% for IntelliJ IDEA Community Edition users, indicating a taste for Scala.
  3. Compared to the average for No Build Tool (4%), this choice is less common among IntelliJ IDEA users (2%) than any other IDE. The only build tool less popular for IntelliJ IDEA users than “none” is actually Ant (+/- Ivy) (7.2%), which drops more than 50% in usage compared to the average (16.5%).

NetBeans (10% IDE market share)


NetBeans remains in a stable third place in the market, using its proximity to Java and Oracle to its advantage when possible, most recently in the same-day launch of NetBeans 8 to fully support Java 8, and continued easy integration with GlassFish as the reference implementation for Java EE 7. Here’s what we see when it comes to NetBeans users:

  1. Similar to Eclipse, Ant (+/- Ivy) usage (19%) by NetBeans users is raised compared to the average (16.5%), as is Maven usage (68.5%) compared to the average (64%). The indication points to greater project maturity with NetBeans users.
  2. Adoption of Gradle (3.5%) and SBT (0.5%) compared to the average for all respondents (11% and 2.5%, respectively) is considerably lower, indicating a lack of affinity for newer tools and alternative JVM languages.
  3. The biggest surprise is the number of NetBeans users who selected No Build Tool (8.5%), which is nearly double the average.

Concluding remarks (TL;DR)


Eclipse users significantly favor Ant (+ Ivy), where usage is highest among all IDE users, and Maven is aligned with the average. But Eclipse developers are less likely to use alternative build tools like Gradle and SBT. This indicates to me that Eclipse represents the vast majority of legacy enterprise projects out there.

IntelliJ IDEA

IntelliJ IDEA developers use Ant (+ Ivy) about half as much as the average, Maven is consistent and usage of newer alternative build tools like Gradle and SBT is nearly double the average. They are also the most likely developers to use any one of the build tools out there. This implies that IntelliJ IDEA users are more cutting edge and less willing to use legacy tools in favor of alternative solutions.


Ant (+ Ivy) and Maven usage among NetBeans developers is slightly higher than the average, which explains the lower numbers for Gradle or SBT. NetBeans users are more likely to ditch the build tool entirely than any other IDE users. This indicates to me that regardless of what NetBeans projects are up to, they need less tooling than other devs (because they got mad skills?).



Responses (13)

  1. Avatar  

    Walter Nyland

    October 14, 2014 @ 3:53 pm

    Note that this doesn’t say anything at all about users of Eclipse, IntelliJ, and NetBeans in general — it just says something about those users that happened to do this survey. Heaps of Java developers in the world don’t use ZeroTurnaround products and hence were not aware that this poll was running and therefore didn’t participate in it.

  2. Avatar  


    October 14, 2014 @ 3:58 pm

    You’re kind of right. But I guess the survey is actually promoted via many different channels, not only among ZT customers.

  3. Avatar  

    Walter Nyland

    October 14, 2014 @ 4:21 pm

    Personally, though I don’t use ZT products, I started the survey, but there were waaay too many questions to answer. Surveys should be much shorter to be effective at all.

  4. Avatar  

    Simon Martinelli

    October 14, 2014 @ 4:43 pm

    NetBeans projects are Ant-based. That means every NetBeans user is using a build tool!

  5. Avatar  

    Walter Nyland

    October 14, 2014 @ 4:54 pm

    Yup. How can NetBeans users “ditch the build tool entirely”? Each project is either based on Ant or on Maven. That’s it, that’s all, except if you’re using Gradle. It’s impossible to create a project in NetBeans without a build tool.

  6. Avatar  


    October 14, 2014 @ 8:42 pm

    Perhaps, those who responded with “Not using a build tool” just don’t realise that what they actually use? That also happens :)

  7. Avatar  

    Cedric Chaveriat

    October 15, 2014 @ 12:54 pm

    I personally think that it could be intersting to compare IDE vs web frameworks
    I’m an Eclipse and Netbeans user, I find Netbeans works well for pure JavaEE applications (JSF + EJB + JPA + WildFly/GlassFish…) while I use Eclipse for almost everything else

    Maybe IDE are specialised in some frameworks. If it’s the case it will be interesting to see what IDE suit to what frameworks to choose the right one

  8. Avatar  

    Dean Schulze

    October 16, 2014 @ 3:43 pm

    I wonder what explains the minuscule adoption of Gradle even though it was released over four years ago. It can’t be because Maven users actually like Maven.

    I’ve tried Gradle a few times and have found that it is un-intuitive, difficult to understand what it is actually doing, and the documentation is not helpful at all to someone trying to migrate from another build tool.

    Maybe SBT will get it right.

  9. Avatar  

    Chris M

    October 17, 2014 @ 8:49 am

    Your Eclipse conclusions aren’t quite correct. It’s not that “Eclipse users significantly favor Ant (+ Ivy)” but that “Ant (+ Ivy) users significantly favor Eclipse”.

  10. Avatar  

    Oliver White

    October 21, 2014 @ 7:39 am

    We tried to keep this survey short at 10 questions, average time to do it was about 100 seconds. We can make a mini version of it if you want!

  11. Avatar  

    Oliver White

    October 21, 2014 @ 7:42 am

    IDEs vs. Web Frameworks is a good idea too–we’ll continue with this style for a while, or at least until the results become too mundane…

  12. Avatar  

    Oliver White

    October 21, 2014 @ 7:43 am

    Roughly 100,000 people were invited to take the survey (about 2.2% actually did), from dozens of countries and via different channels–definitely not just ZT customers! The JUG community was also helpful in sharing this too.

  13. Avatar  


    November 8, 2014 @ 11:29 pm

    Netbeans users “don’t” use build tools more often, because netbeans excels in managing build process, so they don’t look under the hood.

RSS feed for comments on this post.

Leave a comment