How is Drupal 8 doing?

The one big question I get asked over and over these days is: "How is Drupal 8 doing?". It's understandable. Drupal 8 is the first new version of Drupal in five years and represents a significant rethinking of Drupal.

So how is Drupal 8 doing? With less than half a year since Drupal 8 was released, I'm happy to answer: outstanding!

As of late March, Drupal.org counted over 60,000 Drupal 8 sites. Looking back at the first four months of Drupal 7, about 30,000 sites had been counted. In other words, Drupal 8 is being adopted twice as fast as Drupal 7 had been in its first four months following the release.

As we near the six-month mark since releasing Drupal 8, the question "How is Drupal 8 doing?" takes on more urgency for the Drupal community with a stake in its success. For the answer, I can turn to years of experience and say while the number of new Drupal projects typically slows down in the year leading up to the release of a new version; adoption of the newest version takes up to a full year before we see the number of new projects really take off.

Drupal 8 is the middle of an interesting point in its adoption cycle. This is the phase where customers are looking for budgets to pay for migrations. This is the time when people focus on learning Drupal 8 and its new features. This is when the modules that extend and enhance Drupal need to be ported to Drupal 8; and this is the time when Drupal shops and builders are deep in the three to six month sales cycle it takes to sell Drupal 8 projects. This is often a phase of uncertainty but all of this is happening now, and every day there is less and less uncertainty. Based on my past experience, I am confident that Drupal 8 will be adopted at "full-force" by the end of 2016.

A few weeks ago I launched the Drupal 2016 product survey to take pulse of the Drupal community. I plan to talk about the survey results in my DrupalCon keynote in New Orleans on May 10th but in light of this blog post I felt the results to one of the questions is worth sharing and commenting on sooner:

Survey drupal adoption

Over 1,800 people have answered that question so far. People were allowed to pick up to 3 answers for the single question from a list of answers. As you can see in the graph, the top two reasons people say they haven't upgraded to Drupal 8 yet are (1) the fact that they are waiting for contributed modules to become available and (2) they are still learning Drupal 8. The results from the survey confirm what we see every release of Drupal; it takes time for the ecosystem, both the technology and the people, to come along.

Fortunately, many of the most important modules, such as Rules, Pathauto, Metatag, Field Collection, Token, Panels, Services, and Workbench Moderation, have already been ported and tested for Drupal 8. Combined with the fact that many important modules, like Views and CKEditor, moved to core, I believe we are getting really close to being able to build most websites with Drupal 8.

The second reason people cited for not jumping onto Drupal 8 yet was that they are still learning Drupal 8. One of the great strengths of Drupal has long been the willingness of the community to share its knowledge and teach others how to work with Drupal. We need to stay committed to educating builders and developers who are new to Drupal 8, and DrupalCon New Orleans is an excellent opportunity to share expertise and learn about Drupal 8.

What is most exciting to me is that less than 3% answered that they plan to move off Drupal altogether, and therefore won't upgrade at all. Non-response bias aside, that is an incredible number as it means the vast majority of Drupal users plan to eventually upgrade.

Yes, Drupal 8 is a significant rethinking of Drupal from the version we all knew and loved for so long. It will take time for the Drupal community to understand Drupal's new design and capabilities and how to harness that power but I am confident Drupal 8 is the right technology at the right time, and the adoption numbers so far back that up. Expect Drupal 8 adoption to start accelerating.

Comments

David Rothstein (not verified):

"As of late March, Drupal.org counted over 60,000 Drupal 8 sites. Looking back at the first four months of Drupal 7, about 30,000 sites had been counted. In other words, Drupal 8 is being adopted twice as fast as Drupal 7 had been in its first four months following the release."

But the overall Drupal user base is about 3 times bigger today than it was back then. So from a "what percentage of sites are running the new version" standpoint, Drupal 8's adoption is actually slower.

It's too early to tell if that means anything though. It will be interesting to see how adoption goes in the next few months. That's the point in Drupal 7's cycle at which adoption started to skyrocket. Here's a good look at the first 8-9 months after the release of Drupal 7.0: https://web.archive.org/web/20111005021818/http://drupal.org/project/us…

Campbell Vertesi (not verified):

I wonder what the list of decisive modules is at this point. Certainly most critical modules for simple sites are ported already, but the enterprise sites that are such a big strength for Drupal need some more critical integration modules. Search API and Salesforce are notable laggards. D2D Migration is still in "experimental" status.

Perhaps some of the sprint time at NOLA should be focused on this kind of critical enterprise module.

Bill Winett (not verified):

I can tell you that what we are most interested in are Domain Access and migration of multilingual content.

Steve (not verified):

I was happy to stumble upon this blog post today because "How is Drupal 8 doing?" is the very close to the question I was looking for an answer to. More precisely I want to know, Is Drupal 8 mature enough yet for me to start migrating sites to it? I have about 40/60% D6 / D7 sites and am feeling quite a bit of pressure to migrate the 6 sites. I am also reluctant to start any new projects in 7. Your post here does not really answer my question but it seemed encouraging. But then I started wondering if this site is in D8? Builtwith.com says no - D7 - https://builtwith.com/buytaert.net Wondering why you have not made the move yet. Are there some things you are doing here that you cant do with D8 yet?

Dries:

Good point! I can't wait to upgrade my site to Drupal 8 because it will be better than Drupal 7 in many ways. There are no technical blockers. I just need to find time. Between my work on Drupal, my Acquia job, the Drupal Association, my family and other obligations, there simply isn't much time left in the day. Upgrading my site from Drupal 7 to Drupal 8 hasn't been able to bubble up to the top of my priority list yet. Hopefully soon!

Steve (not verified):

Thanks for the response. Wondering: what would your process look like? How much time do you figure it would take you to do? Glad to see all of the comment activity to this post. Learning a lot.

Dries:

My biggest challenge is likely to be the theme; I'm not a theming expert. I'm fairly confident I can upgrade my site in less than a day -- unless I end up fighting with CSS for 5 hours, that is.

Dries:

Drupal.org is a big and complex website that uses a lot of contributed modules. It are typically the simple sites that get upgraded first. Drupal.com runs on Drupal 8, for example.

catch (not verified):

Drupal.org used to run release candidates of core up to and including the release of Drupal 6. It was Drupal 7 where that tradition stopped. While it would not have been remotely feasible to update Drupal.org to a Drupal 8 release candidate, I think we should seriously consider whether it's feasible that it could run a Drupal 9 release candidate when that eventually happens - both in terms of how Drupal 9 is developed and the relative sub-sites of d.o (association? api? even if not the main site).

Dries:

Agreed that would be very valuable. Running Drupal.com on an early beta release of Drupal 8 helped us identify and fix bugs in core.

Many years ago, when I still ran Drupal.org, Drupal.org was always on HEAD. I believe WordPress.com still runs WordPress' HEAD branch. Would love to see more people/organizations do that ...

Fredrik Johansson (not verified):

As far as I know, both Pathauto and Field Collections are in alpha. Even though they might be well-tested and work well (I'm not sure if that's the case), the "alpha" label does not signal that.

Genzo (not verified):

Hi Dries,

The annoying thing about Drupal is its documentation. It's so hard to find your way on it. I think it is needed to be renovated.

hans rossel (not verified):

If I look at https://www.drupal.org/project/usage/drupal, the Drupal 8 adoption has not moved much since 15 november 2015. The most used contrib modules of Drupal 8, Token and Ctools, have each 10000 installs. So I don't agree that the adoption is going faster than for Drupal 7. In Drupal 7 I could launch a site including Drupal Commerce (which has a lot of dependencies) 6 months after the release.

There is also need for a Drupal 8 version or valid alternative for the Webform module. The core contact form has very limited functionalities and eform is in dev.

Mark Hanna (not verified):

+1 on it not moving as fast as Drupal 7. I know our shop adopted D7 quicker than D8. We're only now considering building a new site in D8.

+1 on the porting of Webform. I think it is unfortunate that Acquia decided not to help with Webform. That's probably one of the most popular and useful modules. We use it all the time. If there was one module that the big boys should help get ported, it was that one. I wonder if anyone can explain the reasoning behind that? Seems to me like a form builder and tons of options like Webform and its sub modules have, would be almost as important as Views, and should be considered to get into Core. When I found out that the port for Webform was literally YEARS away, it really cooled the interest in D8. If a major module like that takes years to get ported, how long for all the hundreds of very useful modules to get there?

John Kennedy (not verified):

When we started prioritising modules we worked on grid of high / low priority vs. high / low effort. We wanted to have the maximal impact with the D8 MAP funds and so we had to make some hard decisions. When we spoke to the maintainers, they indicated that it would take 2000 hours to port Webform to D8. According to multiple sources Contact Form + Contact Storage can achieve 80% of the functionality with views. We went round and round on this decision for months, but in the end we felt that porting it wasn't the highest impact that we could have.

With that said, I think it would be fantastic if a group of people stepped up. There are still some great reasons to port Webform. For forms with many fields and multiple pages it is more performant. Porting Webform and Form Builder would be great for site builder capability. It makes sense to have a contrib module dedicated to this problem.

Mark Hanna (not verified):

Are you aware of any good documentation / blog articles that provide some guidance for using Contact Form + Contact Storage? That may help some of us make the transition until Webform for D8 is ready.

Former Drupal Dev (not verified):

The reason is that the creator of Webform module is leading the BackdropCMS.org project which is continuing to build on the D7 codebase, and has implemented many of the same features as D8 without totally rewriting the core. Acquia wants only to support the D8 codebase.

Dries:

Acquia might help port Webforms later but for now we're focused on other modules. This decision has nothing to do with Backdrop. See John Kennedy's comment below for more details: http://dri.es/how-is-drupal-8-doing#comment-125301. In the mean time, hopefully other people/organizations in the community will help port Webforms module (if not already).

Mark Hanna (not verified):

How about the use case where you can train your non-developer/non-techy clients to build forms successfully? With tons of options and a sane interface like webform provides?

How about upgrading from webform to whatever? Is there a path for that use case? I think very much NOT.

Its all well and good if you can code or do some complex site building to get a feature. There are tons of websites in Drupal for organizations with tight budgets, who can't afford for a developer to touch every single thing.

We're in a situation where a lot of our clients are small to medium sized non profit organizations. These people do not have enterprise budgets. What made Drupal attractive to them was that they could have advanced features in their websites, oftentimes simply by having a module found, installed, and configured. No dev time required.

A few of these non-profits are on D6, and happy with it. We tell them they have to upgrade. They ask why? Now a lot of them are in a tight spot. Upgrading to D8 is simply not an option because the contrib module space is not ready. Upgrading to D7 is relatively pointless because it too will be deprecated soon (hopefully not). The small to medium organization is in limbo of what to do, and there are no good answers.

IMHO it was a big mistake to deprecate security support for D6, before D8 was actually ready.

And lets get something straight. Drupal is not just Core, Drupal is Core + Contrib. Drupal core by itself is not good for a complex website UNLESS you have the technical chops / budget to develop a ton of custom code. Case in point, Drupal.org

So if you include the contrib space into what you call Drupal, then D8 is STILL IN ALPHA.

Campbell Vertesi (not verified):

Hi Mark -
Please understand that there is more than one group who uses Drupal. You represent an important audience, and you're right that for that audience, "Drupal" is "Core + Contrib". I'm thinking anyone who mostly has to rely on "clicking things together." Your group NEEDS a solid contrib ecosystem to use the platform. We are counting on these users to understand release cycles with a little more depth than a binary "is it ready or not."

With a major release, you have to target developers first. Developer adoption is what builds the ecosystem which "click it together" users need. Drupal 8 was released when it was "ready" for production use by developers, so that they could get actively working on the functionality which you need. The reality is that developer adoption must precede click-it-together adoption.

We all understand that Drupal 8 isn't ready for "click it together" users with complex sites. Those users belong in a later adoption wave - a few more months at this rate. Depending on the complexity of your site, I would advise still building on Drupal 7 into 2017.

If you're going to be at Drupalcon NOLA, I would love to meet with you at the sprints. Maybe we could help Webform 8 or EForm stable get out the door, with a solid upgrade path.

Jacob Rockowitz (not verified):

The webforms module has close to 500,000 installations (https://www.drupal.org/project/usage/webform) this means there is a huge need and opportunity for new D8 form builder modules to emerge, as well as external form builder solutions to integrate with Drupal. If a contrib module provided better integration with applications like SurveyMonkey or Google Forms, this could provide a simpler and very robust UI for non technical users.

I also see JSON Powered Forms in Drupal (https://events.drupal.org/neworleans2016/sessions/json-powered-forms-dr…) having a lot of potential.

Personally, I am working on a very (very) Drupal developer centric form building solution, called YAML form (https://www.drupal.org/project/yamlform). I am targeting the other 20% (maybe just 10%) of the form building usecases that the Contact Storage module may not be able address.

I worked on an early (pre beta) adoption of D8, and the lack of webforms was the only really 'pickle' of a problem, as long as you are okay applying some patches and occasionally writing some custom code, D8 is an awesome platform.

catch (not verified):

How about the use case where you can train your non-developer/non-techy clients to build forms successfully? With tons of options and a sane interface like webform provides?

The form building UI for webform is from https://www.drupal.org/project/form_builder

There are (inactive) core issues to use form builder or something similar for the field UI, for example https://www.drupal.org/node/1472348

For me personally, I'd love to see work done on that - this would then improve the UX for all entity type creation as well as making contact/contact storage a more viable replacement for your use-case with webform.

Ted Bowman (not verified):

The Drupal 8 core contact form plus Contact Storage will get you a long way to what EntityForm did in Drupal 7. Contact Storage probably could be thought of more as "Contact Plus". It adds more than just storage. https://www.drupal.org/project/contact_storage

I am the maintainer of EntityForm and eForm(the Drupal 8 version). Honestly the reason that eForm development hasn't moved faster is that most of the people that I know that have looked into eForm really only needed what Contact Storage offered. I don't know if eForm should be finished or resources should just be put behind Contact Storage.

There are a couple of developers working on eForm now and I would be willing to help other developers get involved.

Mark Hanna (not verified):

I wonder if there are some statistics about how long it took major/popular contribution modules to get ported from D6 to D7 versus D7 to D8.

One of, if not the greatest strength of Drupal, is the huge base of free, contributed modules.

I sure hope 2 things:

  1. D9 is not such a radical change from D8 that every single module needs a complete rewrite.
  2. That D7 is not deprecated and left without support for at least 3 years.
catch (not verified):

For me personally, the new release cycle makes it much more feasible for 9.x to be a smaller jump - because the transition will be from 8.6.0 or 8.8.0 to 9.0.0, not all the way from 8.0.0. Every previous major branch opened either before or shortly after the first release of the new version, but 9.x won't open for at least another year or so.

There are three issues discussing planning around this. They haven't received a lot of attention because 9.x is rightly a long way off, but I do think trying to get some sort of consensus for how we approach this would be good so that people can begin to plan things now. For example it makes a big difference when considering whether to update from 7.x to 8.x or skip it and wait for 9.x.

https://www.drupal.org/node/2598662
https://www.drupal.org/node/2608496
https://www.drupal.org/node/2608062

Bruce Clark (not verified):

Well I've nearly finished building my first site with Drupal 8 - not quite online yet. I am no whizz kid when it comes to tech stuff but managed it! I used Acquia Desktop 2 and there were teething problems but worth the adventure and great help service. Not using the default themes I struggled with finding a responsive theme that worked although it might have been solved now but Omega theme worked a treat straight away. Thanks for all the hard work to contributors and hope the projecy goes from strength to strength :)

Osmanys Fuentes (not verified):

I think that D7 was the radical change that people was waiting for, but D8 doesn't feel neither so radical neither so waited for. D7 was terrific because of the increase of the easiness compared to D6 and the adoption of new technologies and standards. D8 is doing the same and a little more, but this "little more" is not so urgent or even is not a need for everyone. In my opinion the strategy of D8 is narrow the market and focus on certain specific applications like big enterprise portals, online commerce, media portals and server mobile apps when websites must deal with big amounts of content and traffic, security concerns and user interactions.
Thinking that way is totally normal that the adoption for D8 is slower compared to that for D7. That including to five years of a great product like D7 with a lot of documentation, modules, themes and a huge adoption and the lack of that for D8 are hindering the D8's adoption.

Tim Regester (not verified):

There is a great talk from Drupalcamp London explaining when to go with D8 and when to stick with D6 here https://www.youtube.com/watch?v=mXB2TdpUznU But the last two days I have been stretching it's abilities using mapping and sadly the lack of stable contrib modules is a stumbling block, which is a shame. I want the editing experience but need those maps and geolocation.

That said for building a basic brochure site it must surely be fine with Drupal 8.

Louis B (not verified):

On some simpler sites, the thing that's keeping me from moving forward with Drupal 8 is the lack of ready to use themes, specifically Bootstrap. Drupal 7 Bootstrap allowed me to built a variety of nice sites fairly easily.

Andy (not verified):

Bootstrap is ready to use on 8, we have live sites using it without any issues.

Anoop John (not verified):

60% of the respondents said that they are waiting for the release of the D8 versions of their modules. That is quite a lot. This is a call for all Drupal contributors to put in effort to get all the remaining key contrib modules (maybe top 500 modules) converted to D8.

There was an initiative (D7X?) before the launch of Drupal 7 inviting contributors to convert modules to D7 by the launch of D7. Perhaps there should be something like that the community should organize for D8.

Dennis Foreman (not verified):

I am DESPARATELY trying to find the BASE release for Drupal 8. All I have been able to find are UPDATES to 8 (starting at 8.05), not the full initial core release. Why does a link to the full base not show up on the Downloads page?

catch (not verified):

8.0.5/8.1.0 are full releases. There's no reason anyone should download 8.0.0 and install it - you should always start with the most recent, stable, release.

Nilanjan Ray (not verified):

I tried Drupal-8 Build with my docker with latest version of 8.1, I like change that now it updates the
symfony component along with many more new modules e.g big-pipe.

Personally I believe once the modules like
- ctools
- panels
- page_manager
- Apachesolr
fully port to D8, then D8 is ready to take off to fly for enterprise applications.

julien mary (not verified):

A obstacle to immediate adoption is that all modules are becoming broken between two major releases. Switching from php5 to php7 doesn't do that. And this is because things are turning deprecated before being removed, allowing developers to smoothly adapt their code.

Obviously doing that would have a price in terms of code maintenance complexity but also some benefits to ease smoother modules upgrades and earlier adoptions.

This said, the change between 7 and 8 is so huge that I'm not sure this could even be possible.

Drupal is not only the core, it is a whole ecosystem of modules and starting a project with a new major version requires the whole ecosystem to have upgraded. For Drupal 7, it was worth to have a look 9-12 months after the release, and realistic to start projects with it 18-24 months after, once things have stabilized.

catch (not verified):

The new release cycle for 8.x helps with this. It doesn't fix the problem updating from 6.x/ 7.x to 8.x, but it does mean that sites on 8.x get regular improvements, and that the 9.x branch is neither open nor likely to open for a while. I posted links about about forward planning for the 9.x cycle, there's the potential there to make that less of a jump even for the major version, without compromising on the changes we actually make over time.

PHP (and many other software projects including most frameworks) don't store any user data. Changes like configuration management in 8.x would not have been possible to complete with a backwards compatibility layer, but improvements to configuration management now it's in place should be. So a lot of work over the past 5-6 years has gone into creating and applying APIs which can then be improved without a complete break later on, as well as on defining https://www.drupal.org/core/d8-bc-policy

David Lohmeyer (not verified):

One thing I noticed sometime in the middle of the D7 release in contrib was a seemingly major shift at not releasing big updates to important modules for 6 months - 2 years. It seems commonplace now to need to use the dev branch of most modules to get important bugfixes, which needs to start changing. It sends a message that modules "aren't ready" and when someone uses the green release for a module and encoutners long-fixed bugs it gives a bad impression.

Nick (not verified):

... that is an incredible number as it means the vast majority of Drupal users plan to eventually upgrade.

Isn't this bullshit? According to your survey, 33% are happy with D7, 20% are still evaluating it, while 3% won't be using Drupal at all.

Dries:

It's true that more people might decide not to migrate to Drupal. Until then, 3% remains a very healthy number though.

drupalwordpress (not verified):

Being a long time drupal user(since D5) I would like to feedback.

Drupal is losing ground rapidly to Wordpress, D8 made it much worse. For the first time in about 10 years I'm switching my sites to Wordpress. I have to wonder if Drupal is on the wrong path.

D8 is slower, too complicated and over-engineered. Before I decided to switch to wordpress, I checked Backdrop, unfortunately some D7 modules I needed are not ported to Backdrop yet.

WP wins by its simplicity, and it has a nice Apps that I can easily edit posts via my phone, that's a big deal these days. Not to mention there are way more modules to choose and ready-to-use, it's always a pain to get key plugins sync-up in Drupal before you can upgrade to new Drupal.

Walter (not verified):

We are happy to see that Drupal 8 is going well even when we switch to Wordpress a few months ago. Reason: Drupal 8 doesn't include media management in core. This is not acceptable nowadays, remind me of the discussion about rich text editor in core in Drupal 7.

Mojzi (not verified):

For the intranet we need a working LDAP authentication module just like D7 has. And also Webform which we used for surveys.

Alan (not verified):

Looking at the dev stats for Drupal, drops the actual usage drops to 61,465 (29th June). I find it very unlikely that 40% of real sites will be running a dev version, so it looks like it may be on par with Drupal 7 with sandboxes or maybe even test bots accounting for the rest?

8.2.x-dev 25,845
8.1.x-dev 14,546

The usage of the most installed module are well under core too, which suggests most installs don't have any contrib modules or the stats are still a bit out? In D7, SEO modules have about a 33% share of total installs (i.e. metatags, google analytics).

Drupal: 101,856 (61,465 without dev versions)
Token: 15,072
CTools: 15,027
Admin Toolbar: 12,334
Pathauto: 11,269
Devel: 7,522
Metatag: 6,740
Field Group: 6,546
Layout plugin: 6,436
Google Analytics: 5,836
Libraries API: 4,019

Or core is just awesome ;)