Um eine Plattform für ein erfolgreiches Software-Ökosystem zu schaffen, bedarf es etlicher Vorraussetzung in technischer Hinsicht. In den Bereichen Architektur, Prozesse und Organisation müssen vor dem Hintergrund einer offenen Plattform, die darüber hinaus von Dritten freiwillig genutzt werden soll, eventuell andere Prioritäten gelten, als es in herkömmlichen geschlossenen Systemen der Fall war.
Welche Prioritäten das sind und warum Twitter hier eventuell manches besser hätte machen können, erklärt Slinger Jansen im dritten Teil des IT-Radar Interviews.
Druckversion als PDF herunterladen...
Interview mit Slinger Jansen - Teil III (MP3, ca. 14 MIN)
Das Interview zum Nachlesen:
Ok, let‘s get a bit more technical. Are there any particular characteristics of Software ecosystems concerning architecture, processes, and organisation?
That‘s a very interesting question, and I think for each of the three types of companies (service-based, SaaS-based, and off-premises-based companies) there are some aspects to be mentioned. Let‘s start with architecture, because in my previous life I was a computer scientist, so I feel safe there. Two of the things you really have to architect for are extensibility and robustness, because in a Software ecosystem anything can happen. I mean your partners will be crazy and they will try to control the energy management component on your phone, they will try to take over the camera and put funny pictures in it, and if you allow for it, then it’s OK–I mean innovation comes from this.
I think, for instance, a company like Layar develops augmented reality applications based on your camera, so if you put your phone on and point with it at something, then you get extra information about, for instance, the building or whatever you are pointing at. I find that is a really interesting innovation and something that a company like Apple or Microsoft or Facebook etc. would probably not have come up with on its own. But of course, this puts great strain on the architecture and good extensibility and robustness are the major points here.
One thing that I have noticed, too, is that I am quite willing to say or prepared to say that the iPhone is more robust than the Android ecosystem or Android operating system at this point in time, and I think that‘s a clear decision made even by the platform owners. I’m not saying that it has something to do with bad architecture. I think the fact that the Android operating system is simply more open creates more opportunities for lower quality experience. But of course you do also enable different as well as more extensibility. This is a clear trade-off and those two companies have made different choices concerning that. That is the first point.
Another aspect I find really interesting is the development pace for the community. If you go to quick, then the community can‘t keep up with the pace, but if you go to slow, then they will probably move to another platform that is more innovative. I think that‘s something you really have to take into account. I am not sure if this is architecture or just the development process, but I have actually seen companies drop out of the Microsoft CRM ecosystem because they only had one or two developers, and they couldn‘t keep up with the new versions coming out, which is very interesting to me.
A further aspect is that you must of course architect for openness keeping your proprietary stuff closer and protect it diligently. That is also a very interesting aspect where you want to give your developers or your partners as much freedom as possible while still keeping your main competitive aspects as close as possible, and of course your architecture has to be as feature complete as possible. Therefore, you can‘t just get away with winging it and writing a little bit of incomplete code here and incomplete code there because your developers or the partners in the ecosystem will find it out immediately. And they will prosecute you for it, because they expect the best of the best. So I think your architecture also needs to be as feature complete as possible. That‘s it regarding architecture. I think there are probably many more aspects to point out, but this is what I can come up with quickly.
Then there are processes. I think, I already talked about it a bit. You really can‘t cut corners anymore. With more openness the architecture is exposed almost immediately. Therefore, the software must be resilient, robust, and trustworthy, and the security must be in order. I think you have to be a little bit proud of what you come up with, and your developers have to be proud, because if they are not, then you are in a bit of a pickle because immediately people start making use of or exploit the weaknesses in your architecture.
Furthermore, I think you are now dependent on your community to do requirements engineering for you, whereas previously we would go to the customers and ask „How would you like your phone to function?“ in the case of phone ecosystems. But I also think that the larger phone ecosystems have to step up to the partners and ask „Hey Rovio Games, this Angry Birds was working really great but what would be the next thing you are looking for? What are the new requirements that you are looking for?“
In that sense I think the whole process of software product management changes for the ecosystem view, and this is actually a project that I am working on right now with the students. We try to look on how software product management practices change when it is no longer just a bunch of customers introducing new requirements, but also your partners and niche players and extension builders and plug-in builders. I think that changes the way in which you translate your strategy into requirements, into functional specifications, into code, and I find that really interesting. That was what I think with regard to processes.
The final aspect, I think, is organisation. I think in terms of organisation, life becomes a lot more interesting. I am even secretly hoping that life will become more democratic so that the smaller players–the mass–also get something to say in these huge power houses that are the current software industry, but I am not completely convinced that this will happen.
Nevertheless, with regard to organisation, I mentioned a couple of points or a couple of observations that I find really interesting. Let us take, for instance, a look at the Eclipse Foundation. When I first started thinking about the Eclipse Foundation I thought „Wow, Eclipse is a beautiful platform“, and I thought that there must be hundreds of developers behind it, so when I think of the Eclipse Foundation the first thing I think of is just a whole bunch of developers. Therefore, I assumed that they would have some developers even on the payroll–maybe a couple of hundred–that would be paid from the sponsorships that the Eclipse organisation gets. However, this is completely not true. They instead employ a large bunch of ecosystem coordinators and practically no development personal. I think they have one or two developers, so I really see that the way in which these organisations are designed is completely different.
Now, if we take a closer look on commercial companies, which are familiar to the audience of this particular interview, I think, then governance becomes a little bit different in the view of ecosystems. The community can now sort of make or break your platform so that you must constantly invest in their needs and wishes. As I said previously, all your partners now become first class citizens in your ecosystem, and I think that‘s really relevant.
Moreover, in open source ecosystems governance is done by the community and for the community, which is much more democratic, and I think it makes the joining of such an organisation a lot easier.
I did some research into why people join certain ecosystems, for instance, why do you choose the Microsoft CRM ecosystem as opposed to the Salesforce ecosystem or another CRM system you can think of. Very often I hear of course that the choice is due to business opportunities, but I also hear a larger cry for openness, and people are saying „We want to have more control within the ecosystem, and we don‘t want to develop some kind of really cool niche functionality that Microsoft finds interesting, immediately buys or recreates, and then you are gone.“
For instance, I think Twitter made some very bad moves in that respect. Their ecosystem was doing all kinds of cool stuff, and they were building a really cool picture implementation or integration into Twitter and all these other things like Twitter clients etc. Then Twitter themselves started sort of making some bad moves by purchasing very specific components or very specific niche solutions and saying „From now on, this will be the de facto niche solution, whereas all the competing niche solutions would then be sort of blown away. We have seen this happen in many cases. Now, if you look at investors who invest in Twitter start-ups, you see these investments have decreased recently, and there I think you can truly see that Twitter is making some very bad governance moves. If I was on the board of Twitter, I would probably have made some different choices. I think that is quite a pity.
Furthermore, I think some terms of organisation must be rethought. In your organisation you should start thinking „What would be the functionality of this app store or this platform? How will we perceive of ourselves as a software company? Are we just going to serve end users or are we going to be sort of a middleware layer or platform builder who approaches the market in a different way, who coordinates the ecosystem in a different way, orchestrates it in a different way? Releases have to be coordinated with partners.” I think for all these areas there are some huge changes to be made, and if you are a traditional software company, I think it is going to be a challenge, but also lot of fun. I think working with partners is even more fun than working with customers. You probably know the Ford saying: „If I had asked my customers what they wanted, they would have said faster horses.“, and I think that is a very good saying by Henry Ford. If you look at your partners, they tend to be a little bit smarter and view your platform also in a five year term or maybe even longer term perspective, and they will really have some very good innovations that they can build in your platform. In that sense, I think it is a fun move, and I really recommend that software companies make it.
Here in the Netherlands I do see some companies because I‘m the Software ecosystem guy, and they approach me first. But I also do see Dutch companies moving into that direction with truly no fear. I mean they are more curious than fearful, and I think that is just the right way to go about it.
That sounds like a lot of homework for our listeners. In a research agenda two years ago you demanded that research being concentrated on software ecosystems should be oriented interdisciplinary and include juridical as well as biological concepts. Has there been any progress regarding this?
I am afraid to say no. At the moment it doesn‘t happen enough. I wish there was more in this respect. I have seen maybe one or two papers coming out comparing software ecosystems to biological ecosystems. One thing that I did find really interesting was the size of the keystones which was one of the questions. I think I talked a little bit about it already in the Microsoft CRM context where the keystone maybe makes up 5% of the total ecosystem and the niche players make up 95%. I think in biology you see similar things. In some ecosystems if we look, for instance, at the biological ecosystem of the jaguar or actually the jungle, we see that the jaguar can eat about 126 species and thereby is a true keystone because it is keeping the full ecosystem in balance. So if out of those 126 species one gets out of line, the jaguar will fulfil sort of a cleanup routine and keep the ecosystem stable.
I find that really interesting, especially if you look at body mass–the jaguars are fairly light animals–and compare it to the total body mass of the whole ecosystem. I find that a really interesting comparison and would like to see more in that area. But the comparison with biological ecosystems has a tendency to get very hypothetically and theoretically, too. Therefore, I tend to stay away from it or even shy away from it, because I think „Does it really help us in any way if we can make this comparison? Will it really introduce new insights?“, and I think that are relevant questions, but not for me.
Now, if you look at the juridical typed stuff, I think that there are some subareas that are being covered. So if we look at, for instance, open source and open standards, I think some people are working on this. But to be completely earnest, I find it still is very immature work, and I am really looking forward to more people introducing more work on open standards and on how complete software ecosystems can evolve around the open standards, for instance, about XBRL–the Business Reporting Language. That is what I’m interested in, and I love to know more about that including the legal implications of whatever kind of changes they will bring.
Read more about the future of software ecosystems regarding the current stage of research and the penetration of enterprises in the last part of the interview.
Das IT-Radar-Interview führte Vincent Wolff-Marting.