holch wrote: I would love Limesurvey to compete with professional tools like Confirmit and Questback. But we need to face reality. From what I remember, our Globalpark license probably cost more per year than the yearly Limesurvey budget (good guess, I don't have any exact figures for both).

Globalpark / Questback and Confirmit play undoubtely in the premier league of questionnaire tools. I would be supprised if Limesurvey, a open source tool developed by a handfull of more or less fixed developers, none of which is doing this full time (from what I can see).

Limesurvey can always compete, at least when it comes to pricing

I don't really use other tools much because I know how to achieve nearly everything with LS. Thus, I am not aware of all features other tools offer but from what I can tel there are only some very few feature which are not be available at Limesurvey or can't be implemented easily using JavaScript.

I would even say that with the new Expression Manager features (scoring, complex conditions, ...) Limesurvey is capable of many stuff other tools can't deal with.

holch wrote: Actually I am quite impressed how much you can do with Limesurvey, when you put things in relation.

Don't get me wrong: I feel that there are soo many things that can and should be improved in Limesurvey.

For example the GUI. While it might feel that new and different features are more important, I feel that Limesurvey could gain a lot more users in its own niche (I feel that Survey Monkey and Co are the real competitors to compare with). Limesurvey came a long way in terms of ease of installation and update. I feel this is working pretty well now. But GUI is important and while things work well, it is not the quickest and of course not very attractive compared to the new tendencies in GUI design. Right or wrong, people like to work with beautiful things and a nice, simple, modern and beautiful GUI will impress first time users and testers and motivate them to go forward. But this might be just my impression and I totally get that this would involve major work which might increase the gap in features to professional solutions even more.

I totally agree. At every developer meeting dealing with new features I vote for working on the GUI first. while adjusting/improving a GUI sounds easy, it is really causing us problems for several reasons:

1. Developers like coding new features but they are not that into improving designs
2. We do not have a professional designer available who can contribute the required time to re-design all screens. -> Are there any volunteers!?
3. Remember that the whole admin interface consists of ~30-40 different screens. That takes weeks even for a full time developer!
4. Everyone has a different opinion of how a better design should look like. There have been some approaches in the past to collect example but we never got beyond that.
5. Improving the design isn't just about adjusting some lines of CSS to make things nicer. We are dreaming of having a new drag and drop survey designer (something like that was available at a beta version which was dropped) and doing a bigger refactoring of the whole admin GUI to make it simpler and more user friendly but that requires adjusting the underlying PHP core code which again takes weeks. Not to mention that whole work-flows would ned to be adjusted

holch wrote: But let me tell you something: A lot of questionnaires and surveys go over my desk per year, from the most diverse companies from Europe, the US and Asia. If I would have to guess, Limesurvey could cover 80% of the surveys that are applied out of the box. Another 10-15% will be possible with some kind of T-Partners workarounds in Javascript. The other 5% are either not possible or would need major work on the core. This is an estimate, but I don't think that I am too off.

I think these numbers are a pretty good guess and from my experience of >200 survey projects I can second this.

Often the biggest problem is not the lack of a feature but the stubbornness of the client ("But we have told our customer that we can have text + radio button + radios mixed at one question type!").

holch wrote: When you say that there are almost no alternatives to tools like Sawtooth when it comes to Conjoint, I totally agree. But this gives me also the impression that there is not much market available, or it is a extremely difficult and costly to implement, otherwise another big one would have stepped in.

I have used Limesurvey for conjoint studies successfully. With some additional coding we have been able to connect a DB with conjoint details (product comparison data) with Limesurvey and present the details at a Limesurvey survey. You can find a screenshot at www.limesurvey-consulting.com/integration/

Data analysis was later done using R and special conjoint software.

holch wrote: I am in no way involved or informed about the development of Limesurvey,

There is a developer meeting every Tuesday afternoon European time, everyone who is interested is invited to join us

holch wrote: but I would guess that there is so much small things to fix and keep up to date, that there is little time for "thinking big" and strategic thoughts. We also have to keep in mind what Limesurvey is and how it is organized. This is quite similar for many OS projects.

While they might benefit applying some techniques that are used in corporations, I don't think this is very easy to implement in OS projects that live of the work of voluntaries. As a project leader you can tell your team what to do, if you thing it is the right way. No mather what they think of it. It might not be the best style, but you can. Do that in an open source project and see after a few weeks how many active contributers are left....

You are right. Open Source (OS) projects are organized differently and since every developer is spending his/her free time on this you can't organize this like at an IT company.
We have very limited resources and need to focus on the important things first. That is bug fixing and that already takes a lot of time. There have been > 700 fixes the last 365 days (see bugs.limesurvey.org/summary_page.php for more details).

We have a very long list of new features we want to develop but as said, our resources are limited. From my experience the last years this is something the community can help a lot. Either by contributing new stuff themselves (@those interested please check www.limesurvey.org/manual/How_to_contribute_new_features ) or by sponsoring the development of new features.

Once we had a nice (but no longer maintained) tool for letting users add new features to a global wish list and let others vote for them. That had helped to get an overview of what the community is looking for when it comes to new features. Currently we are scanning the forums and keep an eye on IRC and bugracker to identify such needs but from my point of view that is far from perfect.
So if anyone had an idea of how to better connect community and development team, please share your thoughts.
As most long time users have already noticed we suck at communication


That became quite long now. All in all, I would also love Limesurvey to progress quicker, with a clearer strategy. But to be honest, I am quite impressed and happy where Limesurvey is right now and how it has evolved since I have used it for the first time.

holch wrote: EM was a huge step forward, but we can't expect that to happen all the time. I don't know why TMSWhite left the team, but from what I understood it was because he felt that EM was what he wanted it to be at this point. And he comes back from time to time to the forum to respond...

Tom is an Open Source enthusiast and thus took the time to implement EM at Limesurvey. He used similar features at a custom developed survey software but if I remember correctly it made sense to use a more powerful and flexible tool (=Limesurvey) for his surveys in the long run so he coded EM.
Luckily Tom hasn't left us completely. He shows up at the forums from time to time and helps answering development questions. But unfortunately he is not actively developing EM anymore which is a pity.

Tom had already spent > 1000 hours on this, just so you know what it takes a very experienced developer to code something that fantastic.

holch wrote: What I feel is the most urgent future tendency: Make Limesurvey mobile ready (front end / surveys)! First step: a responsive template among the default templates shipped with Limesurvey.

Currently there are some mobile/smartphone and tablet optimized Limesurvey templates available at www.limesurvey-templates.com/smartphone-...-optimized-c-28.html

I have also seen some approaches for developing a bootstrap based template. Some of these look really nice but most do not yet support all question types.

I'd also love to see some kind of Limesurvey app especially for offline use but technically that is very hard to implement, especially when it comes to automatic data upload.

holch wrote: Then for market research purposes a good reporting would be great. But to be honest, Globalpark's reporting was never very good either (haven't seen it for a while though - talking about 2008 and before).

What reporting feature do you think is missing? The Limesurvey admin statistics support charts, complex filtering and export to different formats (BTW, I think we should improve the documentation, lots of users seem not to be aware of those capabilities).

From my point of view complex data analysis is nothing a survey tool implicitly has to deal with. Powerful tools like SPSS and R are made for this and Limesurvey support exporting data in these formats.

we don't have lime survey in our country..........

Wonder if Mark Zielinski (spacejanitor) is still using Limesurvey in his MR Business.
Seems to have lost interest in the discussion.
The company website mentioned here ( www.dataminingblog.com/guest-post-mark-zielinski/ ) isn't working at the moment.

jelo wrote: Wonder if Mark Zielinski (spacejanitor) is still using Limesurvey in his MR Business.
Seems to have lost interest in the discussion.
The company website mentioned here ( www.dataminingblog.com/guest-post-mark-zielinski/ ) isn't working at the moment.

That page is down since months.

Mark was last seen online 3 here years ago.

To focus on MR is now a bit myopic.

The landscape of survey software is mostly used by users with no relation to MR.

Universities
Students (often onetime-users)
Small- and Midsize-Businesses
Private End-users (often onetime-users)

Developer (using Limesurvey individual survey projects).
SaaS (e.g. Limeservice.com or hostedincanadasurveys.ca )

The vendors of the old, big suites are focusing on big companies. They more and more ignoring universities and market research firms. Market and/or budget them to be to small to be targeted.

The "newcomer" are SurveyGizmo and Qualtrics. Qualtrics is targeting universities heavily.
Just search for Limesurvey and Qualtrics and you will find examples, that it departments if universties are offering Qualtrics instead of LS.

www.limesurvey.org/en/forum/can-i-do-thi...n-favor-of-qualtrics

I thought tight budgets would drive LS into the universities. But it doesn't seem to stick like.
When I compare that to the success of R. SPSS was forced to support R.

I am still trying to figure out what the job offer means for the future of LS?
www.limesurvey.org/en/home/69-limesurvey...r-in-hamburg-germany

For long running surveys Limesurvey with heavy customizing is working. The LS developers own websites illustrate that this is a business.

If IaaS is no option because of restrictions (e.g. data protection laws) Limesurvey is the only option without many strings attached.

But IaaS like LimeService are competing with Qualtrics and SurveyGizmo sooner or later.
techcrunch.com/2013/03/02/the-story-behi...-enterprise-company/

The job offer might be a sign for a bigger plan.

The users won't recognize that hard work and the success of the OSS project "limesurvey".
When I talk to other survey creators (one or two surveys a year) they are seeing many features as a commodity, which LS is missing. E.g. when you restrict a numeric field to a range, that no hint of restriction is shown until the submit button is clicked. Or that error messages are customizable (not via PO-Files edit).

I would like to see templates bundled with Limesurvey, which are offered via subscription.
Browserbugs are fixed and shipped.
github.com/Shnoulle/skeletonquest/commits/master
www.limesurvey-templates.com/improved-st...-templates-c-26.html

It still boils down to the question who collects the money and how is it distributed.

If I put a price tag under each of my feature request, what will happen?
I will offend developers because of low amount in comparison to the hours needed to implement the requested feature.

We could start with three categories 100 / 250 / 500 EUR . Features which are above 500 EUR would have to be split up into smaller requests.

Another source of income (which once seemed to be considered by LS) is the comfort update.
OSS project Contao is using that concept ( update.contao.org/en/ ).
But the market for CMS is a lot bigger than for survey-tools.