Home> Resources> Transcripts> Open Source>

Latest Industry Focus Series:

Open Source as a Business Strategy: Alliances, Marketing and Development in an Open World
[Podcast] [Transcript]

Alliance Management and Certification Workshop
[Info and Schedule]

Negotiation Skills for Alliance Managers Workshop
[Info and Schedule]

When Small Companies Partner with Global Giants: Dating Super Models
[Podcast] [Transcript]

Beyond The Chasm: Profiting From the Next Technology Wave
[Download PDF]


Also See our latest Podcasts, Panel Seminar Transcripts, Whitepapers and Articles

Transcripts - Open Source as Business Strategy

Open Source as a Business Strategy: Alliances, Marketing and Development in an Open World

Presented By: SDForum, Marketing SIG
Moderated by: John Soper, New Paradigms Marketing Group


    Bernard Golden, Chief Executive Officer, Navica, and author of Succeeding with Open Source
    John Bara, VP Marketing of XenSorce
    Bill Soward, President and CEO of Adaptive Planning

(More infomation:

Hear the Audio

Read Abbreviated Highlights

Abbreviated Highlights Also Available

Ed Buckingham (SDForum):

So with that, I would like to take a moment to introduce John Soper who is going to moderate the meeting tonight.

I met John about 10 years ago. He and I have worked together off and on many years back. When we started talking about this, we were figuring out how to make something around the whole Open Source fit the marketing scene. John started an organization called New Paradigms Marketing where alliance management and Open Source are two of the things that he does, as well as what you would call general marketing, business strategy, contract negotiations, alliance development and management.

So, with that, why don't I turn the meeting over to John? John will moderate and we'll have a panel discussion and then, Q&A at the end.

Thank you.

John Soper (Moderator, New Paradigms):

Thank you very much, Ed, and thanks to SDForum for having us here, for all of you coming and to our great panel. So, I'm looking forward to this. I'm excited. I've been involved with Open Source for some time, but I still have a lot to learn, particularly as we move from the infrastructure layer up unto the application layer and it's very interesting to see things work in new and interesting ways.

Hopefully, all of you would want to understand how this can work – what the issues are in making Open Source work and align business development- market approach. So, that's what the focus is tonight. I've been to a lot of panels where we talk about Open Source and for all generic terms, so we'll get a little more focused on the business strategies tonight.

We have a great panel that have some perspectives in a number of different ways, but they've all got a lot of depth on Open Source.

To my left is Bill Soward and he is President and CEO of Adaptive Planning, an Open Source company who will have a lot of interesting things to say. He's also a SaaS company – Software as a Service and I'm particularly interested in seeing how those things play together. Prior to that, he was an Executive in Residence of Accel Partners and a General Manager of FRS business unit of S1 Corporation. He was also CEO of S1 Europe and Edify business unit. Prior to that, he held several executive positions in Edify and actually helped drive the acquisition of Edify in buying S1. Prior to that, he also held a number of management roles with Siemens/ROLM and IBM, and has a business degree in Business Administration from UC Berkeley.

I will let, when we get to his part of the program a little more, I'll ask him to drill down a little bit on his company and what they were doing in the Open Source world.

To his left is John Bara (XenSource) who is a Vice President of Marketing at XenSource. Prior to that, he was the Senior VP of Marketing at Interwoven. Prior to that, he was Vice President of Marketing at Genesys Telecommunications Labs. He was also with Intel and was at a management team for the Pentium group, a Financial Controller and…

You've got a long list of things you did at Intel here.

John Bara (XenSource):

A lot of fun stuff.

John Soper (Moderator, New Paradigms):

A lot of fun stuff.

Also a former Banking Executive with the Bank of Boston and Citibank Tokyo, so a broad background and an education degree from Oberlin which is just down the street from IO Wesleyan, and an MBA from Harvard Business School.

We're also very happy to have with us Bernard Golden who is the Founder and CEO of Navica which is an Open Source consultancy. Bernard has made a business out of being an expert in Open Source. As a matter of fact, granted he is not paying me to do this, he's got a book which I'm sure he'll be signing afterwards – Succeeding with Open Source. So, he's done a lot of speaking, authoring, he's got a blog on CIO online and a lot of consulting, these two being clients…

Is that correct?

…and has had numerous other Open Source companies as his clients.

So, we have some real drilldown depth and some breath to add to the panel with Bernard.

Prior to that, he was a Venture Partner for international venture funds and prior to that, he was Vice President and General Manager of a number of software companies including Informix, Uniplex Software.

So, the way I'd like to approach this tonight is to drilldown on three business case study areas – one being marketing and the Open Source model, how businesses are utilizing that to leverage their marketing efforts. Second, Open Source software – that model and how it interplays with the community-development model. There are a number of ways that can be done – pluses and minuses of different ones, so we'll hear some good perspectives on that. The third area I would like to drilldown on is how Open Source companies play in third-party alliances – anywhere from joint marketing to OEMs to acquisitions such as… Many of you have probably read about the Citrix acquisition of XenSource which we will certainly… We're going to hold that until last. How do you make half-a-billion dollars on an acquisition from an Open Source company? We're going to hear a lot about that in the end, I hope.

So, those are the kinds of things we'd like to drilldown on, but since this is truly somewhat new differently for different people, I'd like to get us all on the same page and Bernard has kindly offered to give us a bit of an overview of where Open Source is today, what some of the challenges are and what some of the leverage points are.

If I can send a microphone there for you…

Bernard Golden (Navica):


So, how many people here work at companies that Open Source is a significant part of their strategy and how many people are at companies that are starting to think about using Open Source as a part of their strategy? How many people came just because they were serving pizza?

Alright, we're going to talk tonight about Open Source software – the new software paradigm and what is interesting about Open Source, what's driving it, what the interesting challenge points are for it and so forth.

Now of course, the first thing I start with is the fountain of all knowledge technical – Dilbert.

Can anybody in the back read this? Okay, nice. This is great eye chart stuff.

So, essentially, the pinheaded boss has found out about Open Source and says, "I want to use Open Source for everything because it's free." Then, we essentially have to reel him in, of course, as they do with everything.

Really, beyond that fact that Dilbert's a fun cartoon and so forth, what this really strikes me is Open Source is hitting the mainstream. Dilbert's making fun of it to an audience that's pretty mainstream. Open Source is getting there.

So with that, let's start off and talk about Open Source.

What is Open Source software?

The reason the elephant's up there is I hark it back to this fable I guess it is, for six blind Indian men who were brought to an elephant and asked, "Please describe it." One grabbed the tail and said, "Oh, it's thick and long. An elephant is like a rope." Another one felt its legs and said, "Oh, they're huge and strong. This must be like a big tree." Another one felt the ears and said, "Oh, this is wavy and flappy. It must be some kind of, like a bat." Then, one felt the tusk and said, "Oh, it's sharp, long and pointed. It must be like a spear." The point of that is that Open Source has a lot of different opinions about what it's about, and many people look at things and say, "This is what it is." But, it's really a conglomeration of a number of things and these are kinds of things that I've drawn out.

One is it's a collaborative development, I think. It's a way for people to work together to develop products and those can be individuals or they can also be companies collaborating. We'll be talking about both of our case studies and frequently XenSource, we'll be talking about how that's interesting, how competitors can work together to create a product that they both take advantage of for their own competitive pursuits – the benefits of that and also, the challenges. But, it is a way for people to work together in contrast to the old way that things were done which is basically, you hired a lot of smart people, you put them in an office at Palo Alto and they build something on their own. This is much more of a worldwide phenomena than anybody can attribute to.

From a perspective of a lot of software companies, it's an inexpensive way to achieve distribution. You put it out there and anybody can download it. We can use it again. You don't have to go tell them about it. You don't have to convince them of it. They find their way to it. They pull it down. They start using it. So, instead of having to go out one-by-one and finding people to use your product, you can really let the Internet distribute the product for you. You can let people start talking about it, so it's a great way to achieve distribution at a lower price.

It is of course, a threat to proprietary software companies although many of them still will say, "Oh no, we don't really see software and outsource to our competitor," or "It's used by small companies," blah-blah-blah. It's a competitive threat and I think that that's becoming more and more clear as time goes on.

It is, of course, a large investment.. Open Source is driven by the licenses that carry it and we obviously talk about that in a couple of slides, but essentially, what makes Open Source Open Source is the license that the software carries.

In contrast to crucial proprietary licenses that are typically custom-done most for every deal where it's kind of, we're going to give you this much usage, this many machines, this many users. Open Source licenses are more or less potential software and say, it's a standard license – these are the conditions you can use it under. The licenses really discipline the way the software can be used.

In particular, Open Source licenses give you a lot more freedom because it's irrespective of the user, you can use it pretty much any way you want to use it. You can go ahead and add as many machines as you want. You can even modify the product because the source code is included and that's a factor of all these Open Source licenses. I'll talk about this a little bit more in terms of the business implications for a vendor. 

Then, according to some people, it is the only software solution on the planet. There are ideologues associated with Open Source for free software because this is the only way that software should be. I mean, there are people in the room who sort of say, "Intellectual property is theft." So, you will run into that, as well.

That's an interesting perspective, but it's all of those things and really, as you conglomerate all those, what it is, it's a huge change to the software industry as we know it – a huge change to the IT industry as we know it.

IEC which is a large analyst firm described Open Source as being the biggest change to the software industry in 25 years, so it's a huge, huge sea change.

Alright, let's put in the next slide.

So, why are companies turning to Open Source? Well, from the vendor perspective, there's the exhaustion of the enterprise business model. It used to be you put those smart people in Palo Alto, they build a product, you get it to 1.0, you need to hire a big, expensive direct-sales force to go bang down doors to get you customers. What's really happened over the last eight to ten years is that model's become exhausted. The buyers stopped buying that way. They got tired of the pace. They got tired of having sales people come and maybe, characterize their products to be more capable than it was. What happened was, from the perspective of the vendors and also from the funders, it just got too expensive to build those kind of companies. You had to pour too much money into that for what you get. So, that model has really gotten very troubling and very exhaustive.

For existing companies on that model like your Oracle or whatever, it's still a pretty good deal, but in terms of a start-up, it's very difficult to try and afford that. So, we need to find something different as a vendor – Open Source.

A lot of reasons that companies look to it is tied to market- and competitive advantages. If you don't have to build something, write it yourself and pay the expense of getting it developed, but you can leverage Open Source that's already out there, you can bring your own product to market a lot more quickly and also you can keep more margin than you would have had in the past. So in the past, if you had to license a component…

I'll just use one of these as an example. If you needed an application server and you bought BEA, you'd be giving up some of your margin to BEA. Given that the market's got tougher, people started saying, "I don't want to give up that much margin," and so they turned to Open Source components as a way of saying, "I can use software, but not have to give up margin – that makes it better for me."

Really, to reiterate the point that I was talking about in the last slide – the ability to achieve distribution and adoption with different time-cost constraints. I mean, you can reach people with your product that you never would have been able to reach before. You can get to them much earlier than you would have been able to before. You can get different geographies that maybe you wouldn't have gotten to for three or five years. Those people can find your product, begin using it, like it a lot, call you up and say, "I want to enter into a business relationship," whereas if you've started with the old model, you never would have gotten your products in and you wouldn't have been able to in-turn your financial transactions. So, it's very attractive from that perspective.

For the perspective of users which is the flipside, why are they interested in Open Source? Why are they willing to use Open Source? Why are they interested in Open Source companies? First and foremost, probably cost. Essentially, it costs a lot less to be going with Open Source. There isn't a big license fee upfront and that's very, very attractive.

It's also because of the lack of lock-in and there's not a of bit coercion – in other words, you've got to give me a lot of money to get access to the bits and the products. Once you've done that, you're locked to me because I'm the company you can get that product, updates and support from. Open Source was not ready for model. The products that are out there – you don't have to pay for that. You can either come to the company for support or not, if you don't want to and if the company doesn't do a good job, you're not locked-in. You can walk away. You can find someone else to support you, so lack of lock-in is a big reason.

Opportunity for customization – I was talking to a pharmaceutical company. This was just a little bird, a bioinformatics company and any real challenge in that, they're big companies, but bioinformatics isn't that large a market and so, the vendors who sell software into that market typically don't have great products because they don't make enough to really invest enough to keep it up to date – put in new functionality. So, they're very frustrated as users.

So, they came to me and said, "What we'd like to do here is put together a consortium of bioinformatics-using companies to build our own products because we feel that we can take Open Source components, customize them and get a better solution for us." So, the opportunity to take that source code and do something with it – very attractive.

Then, the collaboration of the community – Open Source is sort of, inherently associated with the community. The other people are using it and many people were developing it. It's a fountain of wisdom. It's a great way to get information. It's a great way to co-develop and as an end-user, the ability to turn to other people and say, "Gosh, I'm running into this problem with this product," and have them say, "Oh, I had that problem, too – here's how you can solve it," is just a great resource.

Oh I forget. The first time I built an Open Source space system, our group was working on one piece and another group was working with a proprietary product. We ran into problems and we'd post something to a mailing list. Twenty minutes later, we had answers from all over the world. "Here's how you do it, I did it," "Here's some code I did," and so forth. The people using the proprietary product for their part of their project called up the support group and basically had somebody who knew less about the product than they did.

It's a huge difference - the opportunity to collaborate with the community is a great thing.

So, this is what's driving people to begin looking to Open Source.

In terms of building a business strategy around it, what do you have to do? Let's go on to the next slide. Well, one of the things you have to do is ensure that Open Source you use is managed properly and this is true whether you're creating an Open Source product from scratch that you're delivering like a XenSource or an Adaptive Planning, or if you're incorporating Open Source components within your own product. You have to be very certain about the licensing you're using. Remember those licenses I talked about come with certain kinds of conditions and you have to abide by them all. If you don't, typically you're not going to get sued, but somebody is going to come to you and say, "You really need to comply with these license requirements – you either need to change your product to come into compliance with this or you need to remove it," so this is a really big deal. So, meeting compliance with the license is really important.

It's also important because there's a movement around Open Source and you want to rely on the goodwill of that movement. If you're seen as not complying with the licenses, you're going to have problems with your business strategy trying to appeal to those folks.

You have to make sure that your business model aligns with the licenses. If the license calls for the product to be available in a certain way or that something can begin with the product and your business model doesn't align with that, you're going to be in real conflict.

You can use Open Source to get great distribution, to piggyback on it, but if you're not aligned with the license, you're always going to be at cross-purposes and have a really difficult time with your business strategy, not to mention difficulty with your community that's built up around it.

I should say that being Open Source is not enough to guarantee success. I've seen a lot of companies that source and said, "Oh, our product's not doing so well – we'll make it Open Source and that'll solve everything." I kind of call that the Tom Sawyer strategy. It's kind of like, would all of you out there mind coming over and mind painting my fence – would you mind taking care of my product that I don't want to deal with?

Open Source is not enough to make the product successful. You have to do all the things surrounding Open Source to make it successful. I'm talking about that in just a minute here.

You have to make sure that your business model aligns with Open Source realities, but first and foremost, can you build a community? That is absolutely fundamentally crucial to Open Source products being successful. Can you build a pool of people who are using it, involved with it, willing to interact with you, willing to contribute to it and willing to work with other members of the community?

Are you really being transparent? This is a real challenge for many, many companies, particularly companies that say, "I'm now proprietary – I want to go Open Source."

First and foremost, your code's transparent. There's nowhere to hide. You can't put out kind of, junky products and hope that nobody will really know because everybody can read the code, so your engineering group has to become less egotistical or maybe, less defensive about their code because people will look at it and comment on it.

I was working with one company that created an Open Source product for their own hardware product, right? So, they had their own hardware product. They wrote a code to go with it. They turned the code to the main Open Source projects and the person there started to comment back to them and said, "Oh, I've rewritten your code, so it's better," so you see what kind of transparency you want.

But, beyond just the code itself which is clear from the license, there's an expectation that you're going to be more open about things like your product plans. You're going to need more room to engage with people about product plans. The old model of, here, we're showing up. Here's my slide deck. We're going to tell you what the world's going to look like. This shows the dog food you need to be ready to eat. That doesn't work in this kind of a world, so you've got to be a lot more transparent, a lot more willing to engage.

Then finally, your product has to make sense as Open Source. In other words, it has to be something that people want to adopt, download, use, experiment with and contribute, and there are certain products that we can thumb-sense that way and certain ones that don't.

An area that Open Source hasn't done a lot in yet is vertical applications. Those don't seem to have really caught fire.

We've seen applications go Open Source like CRM and stuff like that. That seems like it is maybe getting some traction. Of course, infrastructure makes a ton of sense as Open Source.

So, you've got to have the right business to wraparound Open Source. Otherwise, it's not going to generate the kind of energy, distribution and so forth that you need.

With that, I think that's it.

John Soper (Moderator, New Paradigms):

Thank you very much, Bernard. That's very useful and we will now start to put flesh this out.

First, I want to give John and Bill a chance to get just a little bit of overview on the companies so we understand them, and how Open Source makes sense with them, so that we can put some context.

Bill Soward (Adaptive Planning):

Let me just give a quick overview of Adaptive Planning. We are in business performance management category that in this context, especially with planning, financial reporting – towards operational metrics. If you know Hyperion, a company that was bought several months ago by Oracle for over $3 billion. Hyperion is in that same market, but for enterprise-customers. We are focused on the mid-market.

None of these are my slides, so I'm just talking.

So, we target companies with 100 employees up to 2,000 employees is our focus. Founded in 2003, first customer and production in the year 2004, today, we have about 125 customers. We are a software service company and so initially, our first rollout of our product was as a hosted offerings of a traditional software service model that is, you look at the category that we're in, particularly the mid-market and lower enterprise customers, there's a huge amount of data involved in trying to make those applications work well and so there is more of a bias in that space in favor of one-premise-based alternatives, so it's not just about hosting.

We made a decision back in the spring of 2006 that we would offer an one-premise version of our product in order to maximize our appeal in the marketplace.

I would say that in software service land, the number one marquee company is Salesforce.com and it's about getting thousands of customers. They have 30,000 customers.

Our challenge is business – was how to build business quickly that has thousands of customers generating tens of millions of dollars and do that in a very capital-efficient way.

Salesforce raised $65 million or something at the bubble, NetSuite another famous software as a service company got over $100 million.

We didn't have access to all that cash, so our challenge was how to get big fast and not spend a lot of money getting from here to there or a lot of time.

So, as we looked at it, part one – we need an on-premise solution. Part two – if you're going to introduce an on-premise solution in 2006, what is the absolutely fastest way to do that? There's no question that leveraging Open Source made a heck of a lot of sense.

So, we introduced a downloadable version of our product in August of last year at Linux World in San Francisco and so, we've been out for a little bit over a year. We have over 50,000 downloads in over 80 countries around the world that are taking advantage of our free downloaded express edition product. So, that's now starting to convert into meaningful business for us and it's a meaningful part of our revenue stream now today.

So, our business model is Software as a Service. We believe that the next version of Software as a Service, not the conventional wisdom today perhaps, but where it's going is that as a subscription-based offering – offering that includes software, software enhancements, bug-fixes with all the maintenance and support. In our definition, that is a great business model. It does not require the server to be living in our datacenter that it can live behind the customer's firewall. We can provide our subscription-based servers remotely and connect to technology that's sitting behind the firewall. In fact, the future is, the truth is somewhere in between. Something in the cloud – it's something behind the firewall.

For us then, having an on-premise version of our product, it's also subscription-based. It makes perfect sense. It's very consistent to where our strategy goes.

Ultimately, the core of success for our Software as a Service companies is a very extensive, try-before-you-buy program. For most Software as a Service companies, try-before-you-buy means the 30-day trial. In our case, we've decided to dramatically expand the definition of try-before-you-buy and really, in the Open Source world, it's deploy-before-you-buy. You get to use a free version of the product. It has a substantial amount of capabilities. In our case, you can run a small business or a not very complicated medium-sized business using our free version and never pay us a dollar.

So, we have an extensive set of options for customers. They can download it, try it and use it themselves. We have a hosted version of our free express edition, as well. You can try that. We have trials of our enhanced capabilities, as well, lots of demos on site, videos, all kinds of training – total transparency.

So, the model that's emerging we think, is extensive try-before-you-buy when you're dealing with the mid-market. The dollar size of the transactions is not very high, so you have to figure out a way to close customers in, in a much lower cost of sales and marketing.

The challenge for most software companies here in the audience is how do you reduce sales marketing expense because that's the humongous number that doesn't want to ever go down.

So what we're doing is, we're reengineering the front-end of the sales funnel, trying to offload as much of the discovery and evaluation process onto the customer, and let them do it through a self-service strategy, let them try it, let them work it through, let them see value and then, let them come in and talk to our more expensive sales people on the phone where we decide to close them.

So, the advantage of our model which is a hybrid model Software as a Service as the business model – two choices of deployment; on-premise; on-demand; customer decides; same price for both; extensive try-before-you-buy; let the customers evaluate the product, see the advantages of it; and then, hopefully compress the amount of time that you have expensive sales people talking to them. Then, close them in that way.

Through that whole process, the only thing I'd say is that Software as a Service – the heart of that is subscription-based which means that in our case, customers buy on a perceived basis 12 months in advance. Every year, they'd renew.

So, when you're in a renewal business, it's all about getting the renewal. All the money spent in your run is acquiring the customer. You don't make any money in your run. You make money over time by having very high renewal rates and so, to succeed in that business requires a company to be very customer-centered – in our opinion, total transparency. So, transparency comes in terms of try-before-you-buy. Transparency is here's all of our pricing. Transparency is here's our source code – it's available and you can go look at it yourself. It's all out there, so really, what you're trying to do is match the customer's expectations with your ability to deliver.

Through that process, and having a great sales and support team behind the scenes, you're able to have higher renewal rates.

Our company has well over 90% renewal rates today and we think that transparency and our ability to author all of this to our customers is a good reason why.

So, we're a little bit different. We're also an enterprise application and Open Source. We were the first Software as a Service hosted company in performance management. There are now two today, four years later. We were the first Open Source performance management company and we may be the only one in that category, as well.

So, the enterprise apps in Open Source is becoming more and more mainstream. We're among the first companies to see real success as a result of that.

Do you have a question?

Audience (Q&A):


Yes, just a quick question.

Do you think you could have accomplished the same thing with free, multiple-resource software?

Bill Soward (Adaptive Planning):

A great question. I don't believe everybody heard it – accomplished the same thing with free instead of with Open Source. Highly-debated issue. There was an earlier version of Adaptive Planning that tried a little bit of free and nobody paid any attention.

The biggest challenge for a company that's starting up is getting awareness and people have to find you one way or the other. SourceForge, where we distribute our product today is a global distribution channel that is virtually friction-free and costs us no money, so let me go back to that again. Fifty thousand downloads, over 80 countries around the world in 12 months – it didn't cost us anything to go there. That was a free distribution channel.

Now, we're a company today that's just pushing 50 employees. A year ago, we had 25. How do you end up with a product in 80 countries around the world? What other distribution channel is that available, accessible and free?

For us, SourceForge is a key part of our strategy and the only way you play at SourceForge is with Open Source, so arguably that alone justifies it. But, beyond that in terms of supporting the transparency aspects, what's more transparent than Open Source.

The last piece I'd say is that a key part of our strategy in our market is to have a great partner channel worldwide and get there sooner than later, so to get to those partners, they're finding us. We're not finding them and they're finding us through Open Source and going through that segment, so the answer for us is an emphatic, yes.


Audience (Q&A):


So, how much of your code comes from the community versus from your developers?

Bill Soward (Adaptive Planning):

Yes, so let's look at it differently...I think, if you look at enterprise apps companies, the challenge for the developer community is that you have to have multiple skills. You have to not only be good at Java programming, but you have to understand the domain.

We sell to finance people and so, understanding what finance people want, how that application works and understanding business rules, workflows and all of that isn't for most developers. So, I think it's already more challenging for you to build a traditional, one-at-a-time developer community around enterprise applications. I think it's even more challenging when you're talking about finance apps. That's not the most glamorous place for developers to hang out in their day jobs.

We always viewed our developer communities to be primarily comprised of partners and the people who would get this sooner than later would be our partners and are willing to provide value around our core application, go after their target customer or vertical market. We've seen real traction with them, just spend all day today with the partner who's signing up to localize our product and translate it into one of the Asian languages. Left to our devices, that's at least 12, 18 months away? They're doing it today with their people and are really amazing as with their ability to do a lot of that on their own. For us, the developer community today is primarily a partnership with a couple of customers that are taking the staff out of us once we partner with them.

Audience (Q&A):


But, the core code is mostly developed by you or 100%...

Bill Soward (Adaptive Planning):

Yes, today, virtually all of this is developed by us and in our decision to go to Open Source, the community contribution for code, we always viewed a being a nice upside I would say, in the first couple of years. What we would get a lot sooner was much better customer feedback and we've absolutely seen that around the clock and around the world – it comes in from all kinds of places. Somebody would say, "What about this?" "What about that?" so great feedback and then, to a certain extent, that early response on body identification. Not the fix, but at least, identifying the problem – that's something, so, that's been very helpful to us.

Audience (Q&A):

I think it was…

So, some say, "I want to make sure we get the…"

John Soper (Moderator, New Paradigms):

Sorry, this is just introductory of less than like, two minutes.

So anyway, why don't we hold the rest of the questions and then, we'll come back around. Otherwise, I'm sure i will be able afterwards to answer your questions.

We have a slot at the end of this for questions and as you said, you'd be hanging around afterwards, so let me get John to give up a brief introduction of XenSource here..

John Bara (XenSource):

Yes, that'd be great, sure.

I think I'm going to step over here if you don't mind, just so we can see what we're saying up here.

Sorry, Bernard.

I'm John Bara from XenSource here at Palo Alto. We're the other virtualization company in Palo Alto.

It's very interesting what both Bernard and Bill said today.

So, I think I'm going to start my remarks for those of you that are new to Open Source or evaluating converting to an Open Source model who maybe, have done marketing here in the valley for 5, 10, 15, 20, 50 years – whatever it is.

Everything you just heard Bernard and Bill describe is pretty revolutionary. You might not see it as there yet, but let me just contrast what I was doing three years ago here in the valley.

I was SVP of Marketing at a company called Interwoven which was a $250 million enterprise software company. We had 200 sales people and a lot of system engineers. We had 50 inside sales people. I had 75 marketing people on my team. We spent most of our time buying lists, cold calling, doing campaigns, doing live events and seminars, targeting IT execs. Our average selling price was…I don't know…$250,000. Our sales cycle was 12 to 18 months.

So what I would suggest is, if you're interested in Open Source, all that stuff I just told you about this sort of, enterprise sales model – throw it in the dumpster. It's pretty much worthless. You throw all that crap in the dumpster and you start over when you come to Open Source, and here's why.

In the next five or ten minutes, I'll try to share some of the key learning I had in the last 18 months, some of which were very shocking coming from a $250 million software company, targeting CIOs with $250,000 ASP and BMW-driving sales people who like to make $1 million a year, and so on and so forth, to really bare-knuckle grassroots as both Bernard and Bill described Open Source software sales and marketing. It's really a completely different world. But just like when you study a new language – a foreign language, at first you get into it and it kind of, blows your mind, and you think, I'm never going to get through this.

I remember being in the basement of a library over on college, writing Kanji for hours, hours and hours, and going outside in the snow and saying, "What the hell am I doing here?" It just sort of, clicks in one day and you get it. That's sort of what happens with Open Source, so don't dismay. Don't despair even though you may have spent 5, 10, 15 or 50 years marketing software in the valley, you can teach an old dog new tricks and I'm a firsthand example of that.

So, Succeeding with Open Source – I liked Bernard's title of his book. XenSource…

Next one please, John.

Citrix acquires XenSource for $500 million in cash and stock – that was last month. This was…

What am I doing wrong here?

…so far the high watermark for Open Source software in terms of the sales price and hopefully, somebody in this room can beat that because it seems like a really big number, but I think what you're seeing here is these numbers continue to go up. I think it was JBoss sold to Red Hat last year for $300-$350 million. Now, this $500…

I'm going to change mics here, if you don't mind. How's that? A little better, right?

John Soper (Moderator, New Paradigms):

Way better.

John Bara (XenSource):

Way better, okay.

So, now putting this up to self-promote or congratulate, although it sure does feel good. It's just to show you that you can do, right?

I joined this company 18 months ago. It was in a tank. Bernard came and saw us. He can tell you things weren't that great, but we did a few things. I'll try to show that in the next few slides to try and just put it into a little bit of a cookbook for you.

It's never the same with your company, your market or your segment, but there are a few best practices I'm hoping I can share with you and make it worth your while coming tonight, in addition to the great words from Bernard and Bill.

Next one, please.

Obviously, Citrix saw this as a huge opportunity for them – a very rapidly growing market server desktop application and virtualization. There was a lot of synergy between the two companies and I think together, we're going to go on and do great things, right? I mean, we're a small little company in Palo Alto with 80 employees. They are…I don't know…6,000 employees and 200,000 customers. They run like 90 million desktops worldwide and somehow, they've managed to partner with Microsoft for 18 years and not get killed. We also have a good partnership with Microsoft.

That's all sort of, the future, but talking about how we did it I think, is what matters to you guys.

So, how did we do it? Something that Bernard talked about that I think is significant and also Bill, is there aren't a lot of Open Source companies that come out of nowhere and create a whole new segment.

So, if you're trying to create a whole new segment and use Open Source to do it, God bless you. Come talk to me after and you can share your slide of how you did it with me because I think it's going to be damn hard.

Piling sort of, a risk of a new distribution vehicle over the risk of a new market – it can be done, but it's like two difficult things combined maybe just too much to handle, right – a bridge too far.

So, what we decided to do as Bill and his company have done, as the Linux distros did… Right, what did they do? They picked a known segment – operating systems, right? Operating systems – they said, "Microsoft, Sun, Solaris, HP-UX, IBM AIX – it's too expensive and too difficult – there's a better way." That's Linux, right? Another example, SugarCRM – great company. John Roberts is the CEO – amazing guy, amazing entrepreneur. Known segment with legs – Bill's segment. A known segment with legs – that's a real segment where there's a better way.

In our case, we picked virtualization because that was sort of the impetus of the Zen Project which came out of the University of Cambridge. Our founders were computer science professors at Cambridge and they thought there was a better way to manage distributed virtualized systems. That was the birth of Xen. The Xen Project was really called the Xeno Project which really stands for diversity. Xen, Xeno – that's how the name came up.

Shelve the incumbent – you may want to pick a fight. I'll talk a little bit about how we had great set of founders, one of whom became the sort of, teddy bear with the Open Source community. Bernard talked a lot about the community and nurturing the relations. That's Ian Pratt. The other one who became sort of, our bulldog – our human missile for the commercial side of the market to really be competitive and pick a fight.

What are you really talking about here in Open Source if you're following a known segment with legs is sort of, a David-and-Goliath story, right? The press likes that. Customers like that. Channel partners like that – they're tired of beaten up by a large, dominant player. They might want to listen to somebody new who's going to come along and show them a better, faster, cheaper way.

Create drag – Linux vendors, we got Xen adopted by both Red Hat and the Red Hat Enterprise Linux Five, and Novell and the SUSE Distributions. Also, Intel and AMD – huge sense of rivalry there. They're beating the hell out of each other and leapfrogging each other. If you keep yourself in the middle of something like that where a couple of vendors have a high sense of rivalry and you can show value to both sides, you'll do well. Who knows what that means – EMOMEIMF? Anybody? That's a John Bara special. All my former employees know what that means. As soon as I write that on the board, they go, "Enemy of my enemy is my friend," right? So, enemy of my enemy is my friend – well, 18 months ago when Bernard first came to see us, it was sort of like, VMware, Microsoft and XenSource were the three leaders. Well, what did we do? We forged a collaborative development agreement in partnership with Microsoft. We both decided that maybe together, it was a better approach to competing with VMware. It's kind of funny to think about Microsoft as a minority player in a market, but in this segment, VMware has 95% market share – 95% of a multibillion-dollar market that's growing what percent a year, Bernard?

Bernard Golden (Navica):


John Bara (XenSource):

One hundred percent a year growth – unbelievable! It doesn't go on forever, but it's great right now.

So, CTT – I liked what Bill was saying there. Look at Open Source as a vehicle – that's your lead generation vehicle, not seminars, not tradeshows, not buying lists, not cold calling. Forget all that. Throw that in the dumpster. Think about having a great product that's innovative and distributing it over the web through your community, and making the community extremely active.

The Xen community includes regular contribution from about 300 engineers. How many employees did I say we have? Eighty. Yes, 65 of them are engineers, but what about the rest? The other engineers come from Intel, AMD, HP, IBM, so on and so forth, Red Hat, Novell. How could a company with 80 employees and some nice venture capital build a competitive threat to VMware? One way to do it as Bernard well said, is through Open Source. You're sharing the game by getting developers to contribute from other companies. It's amazing. Guess what? As soon as their products ship, your product may just be optimized and run faster than the competitor's products when Inter ships a new chip, AMD or HP ships a new server, so on and so forth. So, it's aligning your goals like that.

To make fast quarterly releases, we just kept going – boom, boom, boom. Our competitor does pretty much annual releases, so you have an ability to move faster than the incumbent because you are moving with Open Source.

This is when you get into more of the marketing stuff – brand PR websites, Google, watering holes. Okay, make sure that you have a branding strategy. Make sure you own your trademarks because Bernard and Bill both talked about the collaborative nature of Open Source. The flipside of that is it's wild-west, it's very chaotic, there's a lot of pushing and pulling, and if you don't own your marks – you don't assert yourself that this is our project and we own this brand, you can get hijacked very easily.

We had an example where a competitor called Virtual Iron on Boston tried to hijack our brand and did not use our code, at all, and started to tell all the press and analysts that they were actually the leaders of the Zen Project, and that they were a Zen-based initiative. Fortunately, we own the registered trademark. We had a good trademark attorney and we basically shut them down. They had to take it off their website, but it was ugly for awhile. Now, they still sort of, whisper that they're Zen, but we've basically left them in the dust because they couldn't steal our brand.

So, make sure you own your marks.

Audience (Q&A):


Which company was that?

John Bara (XenSource):

Virtual Iron.

John Soper (Moderator, New Paradigms):

When did they do that?

John Bara (XenSource):

About a year ago – April '06.

Bernard Golden (Navica):

They're a former client of ours, so we're kind of surprised that they pulled that.

John Bara (XenSource):

Well, you know. It happens, right?

So, the PR – get a great PR firm. This is a marketing, right? Go for somebody who is going to be who will get you great exposure and PR. Go to the influencers who are not just the writers. Go to the experts like Bernard. He's someone who's a writer that has a lot of great coverage. A lot of the bloggers out there – in our segment, it's like Dan Kuznetsky who used to be with IDC. He's with ZDNet, Charlie Babcock – InformationWeek, Steven Shaglin – Cnet. These are sort of, the go-to people. The people at CRM – they just really work those relationships.

I can tell you that where our firm was – the website should be stellar and it should look bigger than you are. It's your lifeblood. Invest in it. Don't scrimp on it. Make it look great, make it easy and make it so that people can get the product from your website right upfront.

You seeded it through Open Source. Now, what's your business model? I didn't put it up here, but you should know what your business model is. Don't just go blindly and say, "We're Open Source," and then, six months later, you've burned all your cash and you said, "How were we going to make money again?" So know. Are you like a Red Hat – are you going to make money on service and support, even more like a XenSource where we sell proprietary licenses built on top of the Open Source of Zen model or it's something else that you invent? So, know what your business model is upfront.

I can't say enough good things about Google in terms of a marketers great ROI. Those things I talked about before. It used to cost me $50 to get somebody to show up at a seminar and then, an 18-month sale cycle. With Google, I can get somebody to my site for $1, $2 and then, I just have to convert them through putting some breadcrumbs along the way. I'm big on Email forms and sending out newsletters.

I've taken up too much time here, but watering holes – it just means…I had a former employee who taught me that concept…where do these people hang out? Where does your target market hang out? Is it something like the Zigg, is it a developer forum or other conferences where they hang out?

Creating the ecosystem is big. Upsell, upsell – washroom's for peeing. I mean, get it out there. Bill talked about seeding through Open Source, but then what? How do you convert them?

In our case, what we did is we kept a drumbeat going of our company and I actually talked to both MySQL and JBoss. I talked to the people who were in charge of their revenue and their campaigns. I did a lot of benchmarking. I want to encourage you guys to do that. What both those companies told me was they kept a steady stream of communication going to those prospects. They gave a very valuable experience to those prospects.

Bill talked about…I think he said…50,000 downloads. He may never see some of those people, but he doesn't care. What he decided was to give a very functional product.

We did the same thing. We've gotten some huge number of downloads, as well and now, those people are just coming back. They're just buying the product everyday, every week.

So, it's a completely different model. It's more about harvesting than cold calling and that's what I would say about Open Source.

The last slide, if you can put it up, John, is this sort of, this role thing that we have between our two cofounders, both computer science faculty members at University of Cambridge.

The guy on the left is the teddy bear – Mr. Nice Guy, right? So, we kind of did a little good cop-bad cop. He nurtured the community, still does today and will going forward.

The guy on the right, Simon, is more of the lightning rod, right – someone who sort of picks the fight with the VMWare, stirs controversy, keeps the brand going, keeps the focus on you as a company.

It worked really well. I don't think we thought about it going in, but when Bernard asked me to write a few slides about this, it sort of struck me that you might want to think about this type of model because if you just have a founder who's working with the Open Source community and you don't have kind of, a bulldog to assert the company, you may have problems.

So, how does the founder keep all the feelings of all the constituencies considered? How does he keep them all calm? How does he keep them all happy if he's slapping them around all the time?

So, you might want to think about how you have different roles to manage the community. Someone who's the person that keeps things going – the project leader and maybe someone else who's taking a more controversial stance for your company.

That's all I have to say and I'm around for questions at the end.

Sorry it took too much time, but I thought it was kind of, a fresh case study since we just sold the company and I just lived through this for the last year and a half. Thanks very much.

John Soper (Moderator, New Paradigms):

We'll get back to you. Thank you.

We want to cover a couple of points before we open it up. I wanted to go through some of the marketing issues which have already been covered in a number of ways, but I also wanted to drilldown on close development and alliances.

So, let me ask you, Bill… 

So, I'm hearing a lot of executing viral type marketing, global, Internet and so forth. What I'm going to ask you is…and anybody else who wants to chime in…since you're in the 99 percentile of SourceForge, what is it about Open Source that allows you to do that, as opposed to just normal viral Internet Google, etc. marketing that you could do with any kind of software?

Bill Soward (Adaptive Planning):

Let me go back a step before all of you to take all of your marketing collateral and programs, and throw it in the dumpster. They work for certain companies. Some of us were, it's more of a transition and so I will say, we have a very integrated marketing program and as I said earlier, the core of this is test-drive, but we have very extensive lead generation activities. We are still buying lists. We are building a very extensive in-house database. We have narrow-targeting Email campaigns. We do some seminars with some of our partners. We do still show up at certain tradeshows. So, some of the traditional programs actually are still quite useful in our market because our core product – where we came from again was Software as a Service behind the main version. So, when you're selling to people in finance, you have to go to where they are and try to figure them out. For us, the evolution to, they find us, it's all viral and we keep them warm until they're ready to buy and then, upgrade them up to the chargeable version. That is pretty much a work in progress, so we still have to do some of the other programs but will continue to tweak that every quarter.

In that context for us, as I said, SourceForge is a great global distribution channel. It's a lot of free awareness effectively for us and so, for people who go to that place to discover new opportunities for new solutions, they're going to go in, they're going to put in some keyword and they're going to search on it or they're going to use the software they have to assertively drilldown and to find things, so in that context, your ranking at SourceForge for that channel matters. Much like there's Google search optimization, there are things that you would do to try to improve your ranking at SourceForge because it makes it easier for people to discover you.

You don't know who they are. You have to have them register at your website necessarily, but the first part of the process is getting them to find you.

So, a very high percentage of our traffic that comes through is organically created through SourceForge.

A formula for SourceForge ranking is all published by them unlike what Google does, so it's pretty transparent in terms of all the different components that are required to drive ranking, but I think it's clear to us that release early and release often is a key component that drives ranking and drives higher percentile which increases your visibility and increases your viral run-rate, if you will, in terms of what comes out of that channel for us.

John Soper (Moderator, New Paradigms):

Relatedly, what is your attach rate from the people you get out of SourceForge to paying customers?

Bill Soward (Adaptive Planning):

Well, it's interesting again. It's because we came from a hosted world, we keep score in perhaps, a little different way than everybody else does. I mean for us, what we're interested in is paying customers. We're less interested in what flavor they are and so, what we're finding is if you go back, conventional wisdom would be someone would download at SourceForge, install it, deploy it and start to use it. At some point they'd say, "Wow, I want to take advantage of your enhanced capability – I'll install it onsite and manage it myself."

That's sort of Open Source 101. For a lot of companies, that's what they do.

In particular I think, if you're down to the infrastructure level, that makes perfect sense because it's much more of an IT sale.

Now in our case, remember we're selling finance and finance CFO Controller, VP of Planning – those are the people that are signing off on these deals and so for them, the IT organization is a supporting cast member in most cases. That's all what it is.

So, what we're finding is, that yes, some of our customers downloading, deploying and upgrading are staying on premise, but other customers, this is a sales tool, so the ability to download and to discover it – download it to start looking at it and start using it – it's a sales tool. Then, when they identify themselves to us through the different schemes that we have to get them to identify themselves, they say, "Well, what else do you have? You've got a hosted version – how interesting. That sounds a lot easier actually. I don't need to do it myself, is that right?" "Right." So, we're converting downloading customers with the sales tool being the download into paying hosted customers who don't even care at that point about Open Source anymore. What they care about is solving the business problem.

So, our whole approach on this is to give customers lots of different ways to discover the value.  We're totally agnostic in terms of how they deploy it. Our pricing is the same. Our sales people are the same regardless of how they use it.

If I just get them in the door, get them playing with it and then, find the right solution that meets their bigger problem.

So conversion rate then for us… What is the conversion rate when someone uses… First of all, they discover you through the sales tool called the downloaded SourceForge when they buy your hosted version. Give credit to Open Source for that – well, we call it an assist, right? That's what you have because for us, it's the integrated sales and marketing program. We're not religious. We don't describe ourselves as "a commercial Open Source company". If you ask who we are, we're a SaaS company. We offer two choices of deployment and we happen to be extremely open, but that doesn't make us something else.

Audience (Q&A):


Most of the finance people I know don't go on…

Bill Soward (Adaptive Planning):

Yes, they don't go on SourceForge, do you think, right?

Audience (Q&A):


The question is the education piece of educating the customer…

Bill Soward (Adaptive Planning):

Well, it's very interesting is if you know, this is the old world and the new world, exchanging them in like, lightning pasture. Conventional wisdom I think, is the Linux crowd hangs out at SourceForge, etc. and what's interesting to us is in 13 months of experience, a surprisingly high percentage of the people who come through SourceForge have finance titles.

So, a little sleeper story out there is a lot of business people are hanging around SourceForge.

What's that all about? Well, you can download our product and have it installed in an hour. Push one button, click, it would go down and install and you can open it up and you can start working with it and building your budgeting and planning application in our app. You can go a long way these days without having a bunch of IT people. 

So, if Camaru is here from our team, we saw one of the titles was a CFO from the Midwest. So, we can called the lady up, "What the heck are you doing?" She said, "Where are you now? Oh, you're in Silicon Valley. Well, you guys out there you have tons of IT people, I'm sure, in your company. I have actually one IT person." She said, "Well, we don't have those kinds of resources, we're in the Midwest. Out here, we do things ourselves, we actually take care of it ourselves. So, I'm the CFO, I downloaded it myself, I deployed it myself and I'm using it."

So, the future…this is for me web access 1995, for those of you who are old enough to remember that. You have to go hack it together. You bought your kit down at Kepler's. I had an … system, which was a colossal mistake in 1995 for getting web access because there wasn't any network to support it except for Wombat here. 

So, you go through this and it takes you forever to get the thing up and running. Well, that was 1995. Today, it's just there. What if SourceForge and the download world of Open Source application is as 1995, where will we be in two, three or four years? How much easier is it getting? We're creating friction-free package solutions that download and install, like that. Why do you need a bunch of IT people to help you out? That distribution channel is a channel, not just for IT people, but for business people around the world and, hello, it's free.

Audience (Q&A):


In all the Open Source projects that you have seen, download is a good metrics. Even XenSource is a success story. But did you think about an independent company who wants to sustain on its own revenue speed, I will not believe that even XenSource could have survived a strategic acquisition and not like other exit. My question is, I would guess based on my own experience, probably your revenue scheme is less than $2 million, maybe I'm right or wrong. But the question is, how do you grow your company on revenue basis? Not like how many downloads that VCs funded that's fine. Then you can go and say to VC, "I'm on my own. I'm going to be the next Microsoft in real revenues. Not in terms of…

Bill Soward (Adaptive Planning):

So, the answer, in our case is we do the math. To be a relevant company you need…NetSuite is going public with a little over 5,000 customers and they're in the same market size or a little bit under ours, so we'll use that and say, "Okay. Here's an IPO strategy with 5,000 customers." So, how do you get 5,000 customers?

In our case it's subscription base relationships over time. We don't care how we get them. So, to us, Open Source is not a business model, it is a set of techniques that different people use in different ways as Bernard had on the slides. You have a half a dozen different reasons why you might have an Open Source strategy – it's a set of techniques. One of the interesting techniques that we leverage is the distribution channel called SourceForge. Another technique is how to build a partner channel lightning fast, globally because they find you and they can provide value easily around that model. So, those are two of the six things that are on his list that we like. Other people say it's all about developer communities, that was less important. But it's a piece of our puzzle, it's not the only thing that we do.

John Soper (Moderator, New Paradigms):

Speaking of developer community, I wanted to ask John to address a little bit how you've worked out with developers, moving away a little bit from the marketing focus on this to what is the development model. I don't know if it's your quarters, someone in XenSource, who worked with other companies in developing what they refer to as the engine and you put the car together. So, you have a group of companies…tell me if I get this wrong…that are working on the same kind of engine product, then they all go out and they put the car around it. So, different models of Ferrari and the General Motors, and compete with each other. 

As a development model per se, can you address a little bit how that works?

John Bara (XenSource):


John Soper (Moderator, New Paradigms):

If you can find the answers to that.

John Bara (XenSource):

If I find the answers, it will be dangerous. So, in our case, clearly the biggest benefit of being an Open Source company is not the distribution and the downloads, it's the innovation. Again, for a small company, 80 employees to have top engineers from Intel, AMD, IBM, HP, Red Hat, etc., pounding away on this thing and adding new instructions, literally daily, then exposing that code to the world and the community, running open tests and whatnot is extremely valuable.

So that's what, John, alluded to in terms of the Xen engine. Then, any company, since this Xen engine Open Sourced, can take that virtualization engine and build it into their own solution. 

Two examples I cited previously were Red Hat and Novell for SuSE. What's a little bit unique on us is, we then take the Open Source engine and put some proprietary bits on there, especially around Windows, because in the case of Windows, Microsoft won't touch GPL. So, when we did this agreement with Microsoft, basically had to re-implement Xen in a cleaner environment up in Redmond and put some proprietary wrappers around in so that it could work together with Windows. So, we're sort of selling these engines of Xen, we're also selling complete vehicles as our…people like Red Hat. It's a little bit complicated to explain. 

Back to the gentleman's point earlier about customers, we've got almost 1,000 commercial customers, and we started shipping in January. So, that's a pretty healthy rate for one year of shipments. I do agree that downloads are sort of like wind direction, they're an indicator but they aren't the thing that really you should look at. It's kind of like in the dot-com bubble, everyone wanted eyeballs, they wanted to pay to use it or whatnot but it really didn't matter to revenue. You can have everybody looking at your website all day long to see something but they might not have been buying anything. We really use the true indicator, is revenue, customer count and repeat buys. 

So, I can agree with that and I think you'll see a variety of exit paths for Open Source companies. In some case it may be sell, in other case it may be merge. I also see a path to IPO for truly successful companies. We're looking at that as Plan A or Plan B and the other was to come together with somebody like Citrix. If we basically continued on our current course and speed we would have probably done $20-30 million next year – in revenue. We would have been profitable, we would have been on the doorstep of an IPO.

So, those are the choices you look at but I do agree, it's not the download. Those are just an indicator. It's the customer count, the repeat buys and the reference customers, average selling price, size of the deal. What's VMWare's average deal-size? Who knows? It's actually $23,000, and you've got a $20 billion market cap. So, not bad for $23,000 a sale. You make a hell of a lot of $23,000 sales.

John Soper (Moderator, New Paradigms):

The thought in that engine, what's the difference between proprietary companies forming either joint development alliances or forming joint ventures or any form of proprietary alliances?

John Bara (XenSource):

Great question. So, it's probably the level of help we (XenSource) would give those companies. If a company wants to take the Open Source of Xen GPL engine, they take it…we help them. We do assist Red Hat and Novell to some extent, to birth their products but if a company were to do a commercial agreement with us on our OEM products, they would get a lot more specific R&D and an example would be Symantec. We did an OEM agreement with Symantec and our products are being fused together and there's dedicated engineering for that.

Last week we announced an OEM addition targeting the big server vendors, so you see some announcements there where we'll be coming out with the Xen commercial product embedded in some of the major server offerings from the big x86 folks. In that case, they'd probably get a little bit more of a dedicated team as opposed to the typical community, working on the mailing list and whatnot. 

That again is the decision that you have to make. If you're kind of a hybrid company like XenSource, you have to sooner or later, draw the line in terms of how much resource you put on the Open Source community. Where it's essential and it's driving innovation but you might not be getting paid versus I've got somebody like Symantec or Microsoft or Dell, HP, IBM, where they want to do an OEM agreement and financial considerations are exchanged, where do you put your engineers? I think there's a balance point between those two and you may face that as a company. 

John Soper (Moderator, New Paradigms):

Okay. I want to get to your questions but I have one more area I've got to cover and that's third parties, alliances and acquisitions. So, feel free to jump in at any point.

Talk to us about Citrix. How did you do it? Is there something about being an Open Source company that makes it more challenging, that makes it more advantageous? It's not just capitalization.

John Bara (XenSource):

No, actually, I think it's a bit of a misnomer that Open Source companies and proprietary companies can't work together. I think the acquisition of XenSource by Citrix is another signal, as Bernard said in his opening remarks with the Dilbert cartoon, that Open Source is mainstream. Somehow, that company figured out that acquiring a company that has Open Source roots and an Open Source engine was fine. From an IP standpoint, from a valuation standpoint, from a channel standpoint, from a community relations standpoint, they went through all of that.

So, how did it happen? Well, the person who introduced me to the company is a guy called Kevin Compton, who's a prior client at Perkins. Great guy, I've known him for years. He's a former board member at Citrix, so there was a relationship there that helped. Mark Templeton, the CEO of Citrix is…if you haven't met him he is an incredible dynamic person, very hands-on. 

So, he came over to visit us a little over a year ago, as a partner. Just to spend…he had like an extra hour to kill here in Silicon Valley. He ended up staying half a day and spending time with Simon Crosby, our CTO, and myself. He really jumped in right away and he got it, then we didn't hear anything. The whole tone of that discussion had nothing to do with an acquisition, it was more like, tell me about virtualization, tell me about Xen. 

That's how all these conversations start. I think if you go into any of these relationships and saying, "What will you pay for my company?" That's sort of a showstopper right upfront. I think 20 years in this Valley has shown me, grow your business – success will follow. So, I think, you as entrepreneurs, those of you who that have designs on starting Open Source companies should think about what can I do to help my customers solve problems, to provide value to my customers, to my partners and everything else will follow. Don't worry about the price for your company and all that stuff. Worry about delivering value for your customers, your partners and good things will happen.

So, that's sort of what happened in this case. I mean all those conversations go like that.

John Soper (Moderator, New Paradigms):

Nevertheless, the price was overpriced. So, one of the question is, right place at the right time or is there something about the viral effect of it being Open Source that gave that force multiplier effect.

John Bara (XenSource):

Probably a little of both. So, I do a lot of work with the Wall Street types because VMWare is pretty extensive … quiet so they couldn't really talk. So, all the analysts were calling people like Virtual Iron. Mike Grandinetti and I were going to Wall Street, doing all these panels because all of the investors wanted to know what is this virtualization stuff and why is Goldman Sachs trying to get me to pay this outrageous multiple for VMware, whatever. 

So, I spend a lot of time there and if you just look at the multiple that VMware had garnered, and continues to garner, relative to our revenue and our install base is growing so rapidly, $500 million is actually a fair price. Seems a little bit crazy but it is, anyway you cut it, they're getting into a rocket growth market. Don't you think that if there were a real viable alternative to Google, say four years ago, don't you think Microsoft probably should have bought that company? They paid pretty dearly because how many billions have they sunk to try to take Google out and failed? I think Citrix has sort of said, hey, we like this market. It's extremely adjacent to our market and we're going to go after it. 

So, the combined strategy of Citrix and XenSource is about rolling out virtualization across the entire application platform. Server – XenSource; storage – Symantec, who happens to be an enemy of EMC; desktop – Citrix; application – Citrix, and we just happen to all get along with Microsoft and VMWare says the operating systems dead, so, there's some friction there.

John Soper (Moderator, New Paradigms):

So, unless Bernard or Bill has some more comments…

Bill Soward (Adaptive Planning):

I just wanted to say one thing about around alliances, which is for Open Source companies, I think alliances take on more importance than the proprietary alternative. That's because a lot of, with Open Sources, getting known, getting awareness because you don't have a lot of the tools you usually had in the past, where you're basically just hiring a lot of marketing and sales people. Absent that, how do you get known? One of the ways you get known is to make friends in the Open Source world with other Open Source companies. Start integrating with them or something, as a way of getting people who are those company's users were yours.

So, alliances take on, I think, a more significant role with Open Source as somebody to be aware of and pay attention to.

John Soper (Moderator, New Paradigms):

Thank you. I know there are a lot of questions out there so let's dive into some Q&A.

Audience (Q&A):

I have question for Bill, I was wondering why you have the same price for your hosted version and your on-premises version when you might add more cost with the hosted version but you're locked in with the…

Bill Soward (Adaptive Planning):

So, the question was, why do we have the same price for on-premise version as for our hosted version. You would think that hosted version would be more expensive because you have more cost. What's interesting is that we looked at the cost structure for both sides and the multi-tenant infrastructure that we have on our data center, the incremental cost and the next customer isn't anywhere near what you would think it is. So, I think our hosting cost is not so bad. 

In the end, technology living out in the wild behind the firewall, you probably had potentially more technical support questions in terms of operability and so forth. We looked at it and we said it could go one way, it could go the other way. The most important thing is keep it simple – one price, done. Customers don't agonize over why we decided with one price, one price is more than the other, it's one less thing to worry about. You worry about the short sale cycle, so [inaudible].

John Soper (Moderator, New Paradigms):

Next question.

Audience (Q&A):


I'll just throw it on to the panel, whoever would like to respond. 

One of the issues I have heard historically a lot. I was talking about some of the BEA solutions that support Open Source in one fact or another. They have the strategy blended, bringing together proprietary Open Source. That's one issue. One reaction I would get, pretty commonly was, love to use Open Source but our legal department won't let us. They're deathly afraid of someone plopping up out of the woodwork and suing us. How do you address that kind of a concern? It's not unique, I've heard it more than once.

Bill Soward (Adaptive Planning):

While they're thinking, I'll just jump in. I think what's interesting is we're at a GPL 2 at this point. At some point, probably we'll do GPL 3 when things kind of stabilize out there. 

The opportunity with upselling customers is to not only sell them enhanced capabilities but also indemnification. So, it may support your conversion process by making that available. So, if someone has a concern about that then great. Buy our commercial edition, sign those agreements, get protected and we'll take that responsibility.

So, I think it's a legitimate issue for a lot of customers, they kind of come through all this and make their own decisions but JBoss was very successful in targeting wealthy corporate Fortune 500 companies with deep pockets and saying, buy insurance policy from us with the supporting indemnification. That's where a lot of the money cam from was through that. 

Bernard Golden (Navica):

What I would add to that is a couple of things. First of all, that's a way that your license can align with your business strategy to provide you with a value proposition that people want to give you money for Open Source products. One thing is just it's nice when there's an alignment, and there are Open Source companies that essentially have that. It's kind of a…maybe I'll informally say it but GPL's got a time bomb and they'll sell you the defusing mechanism if you give them some money. So, that's one thing. 

The other thing I say about corporate attorneys, kind of saying, "I'm not comfortable with that, I don't want that." That makes a ton of sense for an environment where their aware of it upfront, and that's traditional procurement model. We're interested in buying a product, we're putting together an RFP. They get alerted to an acquisition, to a procurement acquisition.

Open Source in a lot of companies never goes that route. It gets downloaded by somebody off in a corner and what happens is it eventually comes to the attorney as a, "Oh, by the way, do you know that we have…blah…blah…blah…" Then they go, "Oh, my God!" There were a lot of end users that kind of start to say, "We better get more of a policy than a process and a process around this because it's already there.

So, in a way, the attorney is saying, "Oh, no, you can't do that." It is in the classic sense, is kind of somebody blocking the barn door after the horse has bolted because the software is in there. I'm absolutely certain that both these companies have achieved customers who somebody downloaded it, started using it, loved it. "Hey this is great! Look at what we're doing. Oh, my god! What is it about this license, we better call them up and get protection."

Audience (Q&A):


As an extension to Bernard's answer regarding licensing and all you'll have to figure out what licenses of the component that you're using, what if I burned that Open Source product and I'll sell commercially, like you guys have, and I'm using certain Open Source components without any given license, area for the licensing, is that possible? The person who is applying that Open Source component on an Apache is much more commercial than a GPL. Then, I have the GPL everything I've done in that commercial space. Is that right or once you've taken certain components they become your IP and general license.

John Soper (Moderator, New Paradigms):

Can you summarize the question?

Bernard Golden (Navica):

Well, I think the question is sort of if you utilize Open Source components in your product, can their licenses impact what happens to your product, and the answer is, absolutely. That's very much in line with what I'm saying that you're product and your strategy has to be aligned with that license and makes sense with it. If you have the wrong licenses in the components you're going in, you could end up with a real conflict between what the licenses require you to do and what you want to do with your business. So, it's very important to examine those licenses and understand them. 

Audience (Q&A):


But general license change. If your user component download their component, change their license, like an article in there anymore. Is that possible?

Bernard Golden (Navica):

The question is: can someone who has a product released under an Open Source license then decide, yes, I'm going to change the license. An answer to that is, if they hold the copyright, if they hold the rights to that market absolutely, but they can't change what the license that the stuff went out in before. They can't retroact and change the license, otherwise everybody would do that. "Here," Oracle would say, "Here's GPL." We're going to download and come around and say, "Oh, by the way, we changed our minds. Where shall I send the invoice?" So, you can't retroactively change the licenses, but you can perceptively change the license. In other words, from now on this is the end this license, and you absolutely can do that and it's your right to do that. 

My recommendation would be think that through very carefully because it may seem like, "Oh now, I'm going to start harvesting but you're also going to find your community, all of a sudden, is going to be a lot less enthusiastic. They will either go find a different product to use or they will take that Open Source product that you had, fork it and go off on their own direction. All of a sudden, it's kind of like you pick up your ball, say I'm going to play somewhere else. They say, great! We'll just play our own game. 

I think that's something you have to look at very carefully before one sort of says I'm going to mess around with the license. 

John Soper (Moderator, New Paradigms):

We have time for two more questions.

Audience (Q&A):


I'm starting an Open Source company because it aligns very much with what we're trying to do and I'd love to see what are the Top 2 things that you would recommend in starting a business based from Open Source. 

John Bara (XenSource):

First and foremost I'd say, make sure it's a product and a business strategy that makes sense to provide Open Source. In other words, both these products are ones that can align very well with Open Source, particularly, XenSource. I mean it's perfect for Open Source. If you have a new kind of product, it might not align very well. 

First off, given all the realities of Open Source, I'm going to look for a community, I'm going to look for downloads, I'm going to be more willing to be more transparent because that makes sense for the product and strategy I'm going to do. If it does, that's great, if it doesn't then you need to think of something different. Really, the second thing I would really say is get your license right. Make sure you've got the right license underneath the product or products that you're using. That will also be the case if you're incorporating other Open Source products or components into yours.

One thing that's very challenging for companies is what if you bring in eight different components and they all have eight different licenses? You've got eight different licenses you have to comply with in terms of management. Even if there's not issues around so called GPL viral stuff you need to worry about, you still have eight different sets of conditions that you've got to make sure that you adhere to. Some of them are like, you have to mention the product in your marketing and in your documentation, and others are you have to put stuff in the headers of your file. You've got to make sure that you adhere to all those both from a legal and a ethical perspective as well. You don't want to have a product out there that somebody in the community said, "You're not really doing right by that product you're using." So, those are the two things. 

Bill Soward (Adaptive Planning):

Yes. Just adding to that, I guess the assumption is that if you're a for-profit company, then understanding how you think you're going try to make money downstream with what that business model looks like is pretty important. The other approach is just putting it out there and see what happens and then figuring it out from there. There certainly are a number of examples of where they just kind of stumbled upon the right answer, but more likely, you have to find it, to put garnishes. If you're going to be an Open Source, more often than not, you need to be in a big space that already exist when there are well known competitors that already defined the space and you can come in and hack them to death. You don't want to be out there creating a brand new market space. 

But having said that, then how are you going to compete against them and where are you going to make money over time? Comes right back down to the basics because a number of commercial Open Source companies believe that their revenue stream is going to be support only. For many years, they said, "Oh, we're not going to charge for enhancements." There's one software baseline, everybody gets it and they come when they need support. You can imagine doing that if you were an infrastructure play where you have hundreds of thousands, ultimately, millions of downloads, people trying your product and there's thousands of developments. MySQL started that way, right? JBoss certainly was in that realm. 

But if you're something that's a little bit smaller, the numbers don't add up. There aren't enough of them out there to just only do it on support. So, you have to think about, I'm going to make my money by a new software company, I'm going to sell my IP. I'm going to have enhancements, that's our strategy. Do the upgrade for the enhancements that are not part of the baseline product. You have to know that because that then determines how you position your product, your versions of your product, what goes where and what kinds of people you need to have on your team. 

John Bara (XenSource):

I want to add to your remarks, do it for the right reasons. Don't do it because it's a trend, a fad or a gimmick. Do it because you believe it's the best way to take software to market and get innovation into your product line. You have to walk this balance between providing value, leading your community and being…I don't want to say punitive but sometimes you've got to pull out the stick and lead.

So, you can't go too far one way or the other. Let's say you just do everything nice for everybody and you're Mr. Nice Guy. You could get your pocket picked. Somebody could come in who is evil and just push you out of the way and take control of your project, and fork it, as Bernard said. 

On the other hand, if you're just out there trying to make money and self interest is all that matters, the developers in the Open Source community are going to kick you to the curb on day one. The culture that I think Bernard was sort of describing is, it is extremely collaborated and there's some benevolence that goes on, there's some sharing. I mean HP and IBM work together, Intel and AMD. That's the kind of thing that happens here is that you have to be able to nurture innovation in collaborative environment across competition and when the time is right, you've got to step up and lead. You've got to pull on to the reigns because there will be people who say, "Hey, look at what that company is doing. Our business model is failing; I'm going to go take that project over." It happens all the time. You wish it didn't but it's sort of the other side of Open Source. It's easy for anyone to participate, so they do – for good, for bad or whatever.

Audience (Q&A):


Just a comment, I found there are many people around here they don't really understand of what they're talking about, the VCs understand a little bit more of the city dwellers. There's a few out there but they really don't get it.

John Bara (XenSource):

Sure. So, I'll give you six names right now, okay? Kevin Compton - Kleiner Perkins. He's also got his own firm called Radar. These are our board members: Pete Sonsini – NEA; Kevin Efrusy – Accel; Nick Sturiale - Seven Rosen; and there's two more, John Connors at Ignition Partners in Seattle, is a former CFO of Microsoft. Those are our current board members. We did have Peter Fenton, he was at Accel, he is now at Benchmark. He's also outstanding, Peter Fenton, and all of these guys understand Open Source. Clearly, they do. 

So, if you're finding some VCs who don't understand Open Source or don't believe in the model, don't waste your time, because I just gave you six who do.

Bill Soward (Adaptive Planning):

Yes, and the challenge these days is in some cases, their dance card is pretty well full, they've invested on a lot of Open Source companies, and so you have to have something that's going to be quite unique and compelling because they've already checked a lot of the boxes in terms of covering the main categories. So, there's a lot of stuff that's already been ticked off.

John Soper (Moderator, New Paradigms):

I see other questions out there, I think people will stay up here a little bit, so come on up and ask a question. Thank you all, it was a great discussion and I will turn the floor back over to Ed.

©2008 New Paradigms Marketing Group