Episode 16: Liferay – Digital Experience Software with Bryan Cheung


Bryan Cheung is the Co-founder and CEO of Liferay, a leading provider of open source software. Liferay’s digital experience platform is used by companies like Airbus, HPE, and NASA. In this episode, Bryan discusses how Liferay was founded to achieve both business and social goals — and how they stayed true to their mission over nearly two decades in business.



Michael Schwartz: Welcome to episode 16 of Open Source Underdogs, the podcast where the founders of open source software companies help you transform your business model into a successful venture.

Bryan Cheung is a Co-Founder and CEO of Liferay, an open source content management platform that grew from a church basement to a hundred million in annual revenues without ever taking outside capital.

I read an interview with Bryan in 2018, and it was one of the inspirations for this podcast.

I created a short link to the article. You can read it by pulling up the URL: gluu.co/Liferay.

If you’ve been listening to the podcast in order, there are only three companies that we’ve interviewed that haven’t taken venture capital. They are:

1. Canonical
2. NextCloud
3. Liferay

What I really liked about the Liferay story is that it provides an example of a well-executed business that stays in control of their destiny.

I don’t want to give up too much here. Apologies in advance if the audio quality is not 100%, the interview was recorded on speakerphone. But without further ado, let’s cut to the proverbial tape.

Bryan, thanks a lot for joining us today.

Why Bryan Joined Liferay

Michael Schwartz: You mentioned in one of your previous interviews that you were a reluctant CEO – how did you happen to be one of the founders of Liferay?

Bryan Cheung: I call myself a proto millennial, I’m a little bit older than the generation we typically associate with that moniker, but closely enough maybe where it was important for me to live my life meaningfully, and in early 2000s, right before I started Liferay, I was traveling a lot overseas, I was involved in volunteering with the youth group.

I was just thinking about how I wanted to spend the rest of my life and doing it for, you know, some kind of capitalistic enterprise probably wasn’t top-of-mind for me. But when I met my co-founders and was presented with the opportunity at Liferay, it sounded compelling enough that I thought it was worth giving it a try.


Michael Schwartz: So how did Liferay actually get started?

Bryan Cheung: Brian Chan, who is the Chief Software Architect, started the project at the open source project in 2001. He had been working on the software for about three years, and one of his motivations was, he saw a lot of large enterprises having access to quality software that just wasn’t affordable for nonprofits, and so he started to see the open source movement.

And seeing open source software, being available essentially for free, as long as you are willing to put the sweat and work into making it work, he wanted to make enterprise portal software available to nonprofits as well.

So he was doing that for a few years. One of our first users was a corporate in Los Angeles, he was doing some consulting for them, and at some point the project barged enough that we needed more consultants to join. And it was right around that time that he thought maybe this could turn into something bigger, so he approached me and a couple of the other co-founders about basically starting a company at that point.

So in the middle of 2004, we each left our respective jobs and started Liferay. I have been working as a consultant with the Universal Music, one of our other co-founders had been with Accenture, and another one had been I believe a government contractor of some kind. We kind of had these stable things going on, but we decided to take a risk and start the company in 2004.

Michael Schwartz: Originally, you did professional services, and then at some point you pivoted the business into product?

Bryan Cheung: That’s right. So for the first five years or so, it started with the four of us, and then we started hiring friends of friends as co-workers, and we grew the services practice. Because the software was open source, it was freely available to anyone who wanted to download it. It did take some work to get it implemented, and really the nature of the platform was, it was intended to build custom solutions, while also reducing the time and effort for that, by providing a lot of functionality out of the box.

It was nice because we could propose Liferay as a way for a large Enterprise to get, let’s say a customer portal or an intranet that fit really well with their requirements, whether it was the systems that got tied into it or the user experience on the front end, but it would cost a lot less than they would expect because there were no license fees, and also a lot of the functionality that they would need was out of the box, and we just needed to tweak it a little bit.

In the beginning, and even until now, we haven’t had external funding, and so services was a nice way for us to build up the company and fund a business while we figured out our business model.

Then in 2009, we saw a lot of the other open source leaders like JBoss and MySQL, doing sort of a Enterprise Edition approach, where they would have a stable version of the software that had a full enterprise end-of-service life policy, professional support, maintenance, bug fixes, as well as some legal insurances.

And we saw that model working pretty well for some of the other players in the space and we tried it out, and that really was successful for us. We went from being 90% services-driven up until 2008, to be 90% subscription-driven within a couple of years.


Michael Schwartz: You mentioned stability, and that was one of the things that interested me about the previous interview – how do you think the stability of the platform relates to the ability to commercialize it and to scale?

Bryan Cheung: So in your interview with Mike Olson of Cloudera, he mentioned that he had targeted market of large enterprises, financial services, companies, hospitals, insurance companies, and we experienced something similar.

When you think about the nature of enterprise portal, customer portals – that tends to be something that larger enterprises typically want, a lot of those have a longer relationship with the customer. If you’re an insurance company, they have policies that you want to renew every year, and in the stressful event of a claim, you probably want the customer to be able to transact that as quickly as possible.

So with these kinds of large enterprises that used the functionality of Liferay, they really needed that stability, they needed that assurance that the software wasn’t going to change over a long period of time. That whatever they purchased a specific version for, those were the features that we’re going to work, and that we were going to introduce new things that might be innovative or fun, but might disrupt the experience for the customer.

That really was the core value proposition of our subscription, and it’s what’s really allowed us to monetize.

I think, again, Mike Olson was sharing similar lead that they were able to key in on specific enterprise needs, like compliance or large-scale operations that were specific to the needs of large enterprises.

Now, in terms of innovation and new features, we were also able to do two things. One is, because the open source branch continues to be developed after the release of the stable enterprise branch, we are able to continue to release those new features there. But we also introduced the marketplace, so that some of the additional functionality can be delivered via plugins that customers who want to use these features, can do so by installing it.

Maintenance Period

Michael Schwartz: Red Hat gives a really long, 5-year lifetime product, but we found it’s really hard to maintain those old versions – how long is it stable?

Bryan Cheung: Typically, for each major version that gets released as part of the subscription, we have at least four years of maintenance and updates. Again, these are just bug fixes, and stability, and security updates. And we will continue to support the product even after that 4-year period.

I believe the total support period is about 5 to 7 years or so, to the point where it does go into limited support, where it’s really just more our team advising the customer about known issues and how to work around them, but it is a pretty long life cycle.

Portal V.CMS V. DXP V

Michael Schwartz: We’re getting back to the business, it originally started as part of a portal CMS platform, but the offering has become more expensive so now you describe it as a digital experience platform. Can you talk about how and why that change came about?

Bryan Cheung: You know, one of the challenges we’ve always had with Liferay was, we came from the portal space, we were also known for our CMS capabilities. But those two monikers were a little bit limiting.

When you think of CMS, you think, “Oh, you guys are just like WordPress, Drupal.” And then when you say the word “portal”, people think, “Oh, isn’t that from like 1995?”

And Liferay, a lot of our customers would say things like, “Oh, well, you guys are definitely more than just a portal or CMS.” Or they’d say, “Well, we use Liferay when we need a portal, but we don’t want to use a traditional portal.”

And so, what do I mean by all that? Well, when you think back to how I was describing Liferay earlier, there are sort of three things that people need to Liferay for. One was, they had a lot of very specific business needs. Whether that was systems to integrate with, or customer requirements that needed to have some extra development. But they also needed to quickly build a solution that often used standard functionality and features that would be things like content management, document management, collaboration features, workflow, and forms.

And then they needed to wrap that all up in a really great user experience on the front end. Of course, that started with just web, but extended later to mobile and different connected devices.

So when you call Liferay a portal or CMS, you don’t really fully capture that full range of capabilities, which is, you know, you can get this solution built really quickly, but it’s flexible, customizable, especially for large enterprises, especially for B2B and B2C industries that have that long customer relationship.

We landed on the term “digital experience platform” to try to capture all of those nuances.

I’ll probably be the first one to admit that it’s kind of an industry insider term, no one’s really looking for a DXP, they’re not probably going to google for it unless they heard it from us, or from a competitor of ours, or from one of the analysts.

It’s not the greatest theme I would say, but we wanted to at least try to get out of some of the preconceptions people had about terms like Portal and CMS, and it seems to be doing the job for now.

Value Prop

Michael Schwartz: What would you say is the core value proposition of the Liferay platform today?

Bryan Cheung: It really is connecting companies to their customers through a great user experience, being flexible enough to address those custom business requirements for a large enterprise, while all being easy to maintain, and to have a low cost to market, and flexibility, and innovation for future dates.

Enterprise-Only Features

Michael Schwartz: Are there different features for the enterprise and open source platforms, or is it just a different licensing sort of packaging different?

Bryan Cheung: So we do have some additional features around the enterprise subscription.

One of those is our integration with Elasticsearch. Elasticsearch obviously is something that larger enterprises tend to need, and so we do have an option for customers to purchase that. And then, of course, it is the maintenance and support services provided by our support organization.

One of the things that we hear over and over again from our customers is that we have a world-class support group. Tthey really go above and beyond to resolve issues for them. And more and more we’re also moving toward figuring out how to make our customers successful.

I think in the early days of open source, it really was about, “Hey, here’s this thing that I used to have to pay for that I get for free.”, and that’s really cool. Now, I think a lot of people realize that technology is complex, and having access to the software is just really step one.

How do you actually make the most of it, how do you implement it correctly so that you get the most from it, how do you avoid making some of the mistakes that others have made before.

And maybe most importantly beyond just the software. If I’m trying to engage customers through my site, or if I’m trying to more efficiently manage my business operations through digital means, or if I’m trying to aggregate all the knowledge I have in a company and find it quickly, there are probably some things I need to do, organizationally. Cultural changes, business processes I have to implement alongside the solution in order for it to really deliver the value that I’m looking for. And how can I partner with an organization like Liferay, to understand what those things are, so that I don’t waste my time figuring it out myself.

We really want to build that part of our organization out as well to complement sort of the break-fix support that we already have, and some of the enterprise-only features.


Michael Schwartz: You didn’t talk a lot about license, so you mentioned some extra features that are in, let’s say, the Enterprise distribution, but is there a license difference between the Enterprise and Open Source releases?

Bryan Cheung: Our open source release is licensed with the LGPL, fairly common open source license. I think the conversation around licensing is probably not as heated or ideological as it was maybe about 10 years ago.

When we started we were with the MIT license, which was very similar to BSD. It was a very permissive license, and then we moved to LGPL at some point. One of the things that’s been part of our journey has been figuring out our open source business model.

I think as you know we are self-funded, and so one of the ways that we tried to balance the needs of the community and our commercialization was to move from MIT to LGPL.

At the end of the day, I don’t know if that matters all that much, I think especially with the plethora of open source software out there right now, the most important thing probably is adoption and winning the hearts and minds of developers.

On the Enterprise subscription, we do offer a commercial license, what we found is that large Enterprises that tend to subscribe to our offering understand commercial licensing better, their procurement departments are more comfortable working with that. And we can provide some assurances around that license that might be harder to do with the open source side.


Michael Schwartz: What did LGPL offer that MIT license didn’t?

Bryan Cheung: At the time we made that decision, there were a couple of large companies that were using Liferay as a platform to build different offerings.

One was a collaboration solution, and the other I believe sort of a verticalized customer portal solution for the automotive industry.

We had a feeling at the time that maybe it wasn’t “fair” for these larger companies to take our software and monetize it, without any relationship back to our ecosystem. So we felt that the LGPL might give us some protection from that in the future.

I think in the hindsight, it was probably difficult for those organizations to make those efforts successful without our involvement. I do think ultimately what they tried to do didn’t quite take off.

It’s kind of one of the pros and cons of open source, you get access to all of this powerful software, really without any restrictions but how do you make the most of it? How do you navigate it, how do you develop and customize it in a way that doesn’t get you down a rabbit hole later, where it’s hard to upgrade, or it’s hard for you to take advantage of the innovations and additional features from the mainland community.

I think now that we as an industry overall have a lot more experience with open source. I don’t think some of those challenges we faced back then would be as much of an issue today.


Michael Schwartz: One of the challenges I think for every organization is sales and bringing new customers. What are the primary sales channels for Liferay? Do people just find you on the web, or would you say the partner channel or direct sales?

Bryan Cheung: I would say, in the first 6-7 years, open source really helped us go to market.

One of our penetrations into a large Enterprise was developer-led. You had Java developers at these large banks and insurance companies, who were maybe tired of using some of our competitors who are maybe too complex, and projects built on them might take far too long to get to production.

Liferay was using pretty standard Enterprise architectural practices or the time we were using EJBs at some point and moved to Spring and Hibernate, which were widely accepted, we were compatible with Tomcat, MySQL, JBoss, which were popular at the time.

And so that really got us into a lot of our initial customers as well as in our expansion period that they continued to help us.

With that groundswell of support from developers that got us into a large enterprise, we then started to get recognition, both in the industries that we served, but also with some of the analyst firms like Gartner and Forrester. And then that sort of event gave us the peer recognition and the analyst recognition that gave confidence to some of the mainstream companies to adopt Liferay, because we had to track record.

What then happened, I would say, is that there were two kinds of system integrators. One was smaller maybe boutique integrators that’s had a track record of taking open source software, understanding it, and then implementing it.

Again, MySQL, certainly JBoss and Tomcat, later Liferay, Pentaho, maybe even MuleSoft, and the smaller integrators found that they could provide more competitive bids for building solutions by having open source foundation. And then the larger integrators, companies like Accenture, Wipro, Cognizant – these came later and saw really the strategic value of open source. That the flexibility, the security may be the way that you could better build best-of-breed solutions for complex, digital transformation problems, was an advantage for their customers. And so they started to engage with Liferay and others for some of their strategic customers.

And so that kind of was a very organic process for us, it wasn’t a strategy that we had thought up beforehand, but it’s worked out really well for us.

Startup Marketing Advice

Michael Schwartz: I’m curious if you had any insights to what are some of the more tactical marketing investment you can make as a startup.

Bryan Cheung: Thinking back on our history, so again, following that arc of that early grass roots adoption by developers – a lot of our early investments were in events like JavaOne.

I remember our first JavaOne, one of the things we did was we bought baseball jerseys that had sort of Liferay scripted on the back. And then instead of the player names, we had things like Ajax and JSR.68, which were kind of the things that developers cared about back in the day. I’m probably dating myself a little bit with some of these terms, but we had an ideological discussion about this like, “Hey, no one’s going to take us seriously if we’re wearing these baseball jerseys. “

If we mean to be competing with IBM and Oracle, these are the people that companies take seriously. And then the ones I think that really understood developer’s hearts and minds were like, “No, this is our way of saying we are not like Oracle and IBM.” And we needed to apply our strengths.

We invested in the booth at JavaOne, but instead of having expensive giveaways and kind of large gimmicks, we did the baseball jersey thing. And people noticed, and then they would start a conversation with us, and once they started talking to us about the technology, that’s what got them really excited. And I think we were able to get people to be invested in the community and just start evangelizing Liferay back into our organization.

In a way, we played up our outsider or maverick position, especially for the developers who didn’t want to go with the status quo. And what we found is that those tend to be the most passionate people, the ones who were really going to go above and beyond to sell your brand and your company to their stakeholders. So that really worked for us, certainly in the beginning.

Now in 2018-2019, I think JavaOne maybe is part of the establishment, so I don’t know if that specifically would work but I think the principles would still apply.

DXP Cloud

Michael Schwartz: Is there a cloud-hosted version of Liferay?

Bryan Cheung: Yes, we introduced Liferay DXP cloud earlier this year, and it’s exactly that.

It’s a cloud-hosted version of Liferay, but it’s actually more than just a place to put your Liferay-based solution. DXP Cloud actually provides a full Enterprise pass for managing the life cycle of your software project, from build to deploy to managing our production.

It has a full set of tools and frameworks to do everything, from continuous deployment integration to testing to auto-scaling, to an auto backup service.

It is built to work on top of AWS, and will be supporting other cloud infrastructure in the future, but it’s got another layer on there so you are not just doing AWS but you have all of that functionality there.

It’s really nice, so our team what we found is, when we were starting to deploy complex projects based on Liferay to production, there were all these things we had to figure out. How do you configure Jenkins, how do you deal with Kubernetes. How do you deploy Liferay to Docker and manage those containers, get them all talking to each other, how do you deal with elastic scaling when demand goes up.

So instead of having our customers figure out all those things on their own, we decided to figure those things out for them, and really save them time, effort and money doing that.

We really think DXP cloud is going to be a time saver, and again, on this theme of how do you deliver value to large enterprises. Large enterprises, they don’t want to have to deal with the DevOps aspects of their solution, they just want to get to market quickly so that they can engage customers faster, learn what the customers want, and then evolve quickly to meet those needs.

Michael Schwartz: I realize it’s early, but how is it going in terms of uptake from customers on the cloud offering?

Bryan Cheung: I was actually really surprised because we introduced three or four different products this year, and DXP Cloud, I would say, had really strong interest, maybe the most of all of those products. We actually already have a pretty significant pipeline of opportunities for people who want to use DXP Cloud, so it’s definitely been really successful.

Advice For Bootstrappers

Michael Schwartz: Do you have any advice for entrepreneurs, specifically who want to grow the business organically?

Bryan Cheung: Don’t do it. I’m just kidding. Definitely do it, but it’s a lot of work, and it takes a lot of sacrifice. But I do think the rewards make it very, very worthwhile.

I think one thing is that you have to be very clear with yourself and with your team on what the goals of the business are.

I think for us, I shared a little bit about my background, I never had ambitions to have a position at a company, or to make a lot of money, but I did want to live life meaningfully, and to use it in a way that would benefit others.

And I think that sort of motivation really is shared by a lot of people at Liferay, and we are able to do that through Liferay I think because we have control over how we grow the business, how quickly we grow, and how we balance the needs of the financial success of the company, and the non-financial success factors.

And because we have that strong vision and purpose, I think it makes the sacrifices required of being self-funded worthwhile and meaningful. And what I found is that the people who work at Liferay, who believe in what we’re doing, they are able to get behind that, and they’re willing to make those sacrifices as well.

We have some brilliant people at Liferay, a lot of them work really, really hard, they’re super talented, they can probably be much more financially successful in many other places. You know, maybe you would argue that that would be true of the founders as well, but there is a sense that we’re all in this for something greater than ourselves. And that’s what really holds it together.

So I think, you know, if we were to change that, or if we were to engage with an outside investment firm of some kind, I think we would lose a huge part of who we are and the value we have as a company.

Liferay Non-Profit

Michael Schwartz: I read in one of the interviews that there was a nonprofit associated with Liferay. Can you maybe just come in a little bit about how and when you were able to do that, and what’s been the experience in balancing, I guess, the goals of the nonprofit and the goals of the business?

Bryan Cheung: We are a for-profit company. From our inception, we have used some of our earnings to support nonprofits of various kinds. A lot of it is around problems that we believe can’t be solved in for-profit ways, things like disaster relief or fighting against human trafficking.

I personally think that a lot of people actually don’t want a handout. When I talk to, let’s say, young folks who have been incarcerated and come out of prison, or when I talk to people who live in developing areas of the world that are maybe economically depressed… What they don’t want is some Western entity, whether it’s for-profit or nonprofit, to come in and just build them a school or build them a well, or give them food and clothing. They actually want to find a job, and they want to earn their own money, they want to develop their skills, and help themselves.

One of the things that made it possible for me to get behind Liferay emotionally was this realization that for-profit companies can actually do a lot of good.

It’s good for people to work hard and to feel that sense of accomplishment from building something meaningful – that’s why we believe in so much of what we do as a for-profit business. But, again, you have to complement that sometimes with non-profit means because some of the very big problems are just really hard to solve, let alone do it in some kind of economically self-sustaining way.

Our vision is that the two sides of Liferay can complement each other. I would love to see that play out in a local community. So are there ways where the places where we have offices and do business are complemented by non-profit activities that also served that local community. I don’t think we’ve gotten the formula right yet, but we have some inklings of it.

One of the things we’ve done, for example down in Brazil, we have a very successful for-profit business. They’re members of our open source community decided to join Liferay and start Liferay Latin America, but one of our most talented developers down there came out of the public school system, and in Brazil, the public school system is not as well funded, and even a lot of the middle-class, they’re going to private schools.

Unlike in USA, here I think you can’t find public schools that are very good, and most people do still go to public schools. In Brazil though, being from public school means that maybe you didn’t have as many opportunities as some of your peers. So this developer, he was telling the story that one of his mentors, when he was growing up, was a developer and show them how to write code, in the sense of empowerment and possibility that he experienced when he wrote his first software program was really life-changing.

And so even though he came from sort of this more challenging background, he was able to follow this session around software development and build a career. He has since gone back to the public school where he grow up, and he has organized funding for a library. But he’s also been able to get some of his co-workers from the Liferay office, who were also developers, to go back to that school and teach these kids and expose them to coding and software development.

It really changes the trajectory of their lives, so boys and girls who maybe otherwise wouldn’t see those role models and maybe who had seen just soccer players and kind of dreamt about doing that, now they’re kind of seeing this new world of possibilities.

And, first of all, that is not really a for-profit activity, but I think even on the nonprofit side, I wouldn’t say that it’s required a lot of investment, but really is just people taking their time out and building relationships with that school.

So that’s the kind of impact we want to make, we want to be a for-profit company that engages the community, maybe identify some of the ways that money can help kick-start or catalyze some kind of change in that community, and then hopefully there are ways for those efforts to sustain themselves in the future.


Michael Schwartz: Awesome. Bryan, congratulations on all the success at Liferay, and all the commitment over the years. Thank you for your hard work, and thank you for being a guest on the show.

Bryan Cheung: Thanks very much.

Michael Schwartz: Special thanks to Yotam from Liferay, for coordinating the interview.

Transcription and episode audio can be found on opensourceunderdogs.com.

Music from Broke For Free by Chris Zabriskie and Lee Rosevere.

Production assistance and transcription by Natalie Lowe. Operational Support from William Lowe.

Follow us on Twitter. Our handle is fosspodcast.

Next week, stay tuned for the founders of TimescaleDB.

Popular Episodes

Subscribe to our newsletter
for news and updates