Discover Performance

A former Forrester analyst sees three issues for apps leaders in 2013: Agile, PaaS, and whether the CIO will still be your boss.

While 2012 was a year of change for leaders of apps organizations, 2013 promises even more change. Some of the changes may be small, but their implications could be pretty big. We asked Dave West, chief product officer at Tasktop Technologies and former research director at Forrester Research, for his thoughts.
 
West listed three things to watch for: a trend toward business alignment, the need to explore the potential of platform as a service (PaaS), and Agile’s evolution from dogma to a realistic approach to software development.

Q: What are some of the major issues that apps VPs will face in 2013?

Dave West: VPs of apps have to think a lot about one very interesting problem: where they report into. I think that, increasingly, we’re seeing that a lot of application development is being done outside of the CIO’s organization in the business. Mobile is where we see it the most, but it is also true in other parts, particularly in analytics and big data. Increasingly, the software developers are popping up inside the marketing department or supply chain or whatever business unit.
 
This is part of a larger trend toward the apps group delivering on business goals, not straight technology goals. The CIO’s organization is almost being looked at just to provide the infrastructure and to support that. If they can’t provide it, then these groups look to other organizations like Amazon or HP to provide that infrastructure in the cloud or a virtualized environment. I think most VPs of apps in 2013 and the years after that will be increasingly more aligned with business.

Q: What’s driving that movement?

DW: Mobile is at the heart of it. They say that, in a lot of organizations, mobile will be the primary way you consume applications in the next ten years, so 2013 will be a bit more, 2014 a bit more. Today, most mobile apps are built inside the business. As you build more mobile applications, then it’s likely that more and more of the cool IT spend will go outside of traditional IT organizations. And the golden rule is the guy with the gold sets the rules—which leads us to a world where the majority of application development is being managed by the business in the context of business demands and projects.
 
In terms of why this is happening, I think you have to look to speed and managing the unknown. Mobile apps need to be released very fast, and they solve problems that we are only just learning about. This means that traditional, cost-oriented, very efficient software delivery groups under the CIO can’t compete. They are built to solve well-defined problems with well-defined solutions, and mobile is definitely not one of those.

Q: How can apps VPs prepare for this new reality?

DW: I think the biggest challenge of this kind of decentralization is that you lose the opportunity for control, visibility, consistency, but you improve delivery speed and the value that you give to the customer, and you have a tighter relationship with that customer.
 
If the problem you’re trying to solve would benefit from rapidly delivering something that is very innovative that requires a lot of customer feedback, you should probably look to a more decentralized model. If the problem you are solving is simpler, then maybe a more controlled and centralized process is appropriate. What I found when I was an analyst at Forrester was that organizations wanted to zero in on one model, and that isn’t right. You have to combine your models. Decentralization is here to stay. Centralization is here to stay. Neither is right. Neither is wrong. Manage appropriately.

Q: You say platform as a service is on the rise in 2013. What can we expect?

DW: We’re starting to see more of it. As an app professional, you have to wonder if it’s going to be the platform wars all over again, or is it something a little bit different? PaaS definitely creates a lot of benefits. For one, the ability to rapidly assemble applications from a collection of services that’s already provided and then deploy that application without having to think about the platform it’s running on. PaaS takes away lots of complexity by instantly providing the service in the context of a runnable environment. No more need to debate deployment architectures or make last-minute decisions about database access.
 
This simplicity is the motivation for its adoption. As software becomes more important, and as the business is in the driving seat, then having to rely on expensive software engineers—an increasingly costly commodity—is problematic. The cloud with SaaS and now PaaS provides quicker, cheaper models for building next-generation applications. Of course that comes with a loss of flexibility, and a dependency on a platform vendor, but the speed and simplicity might make it worth it.

Q: Should apps VPs be diving into PaaS or should they be more cautious?

DW: I think they have to decide whether they want improved simplicity. With PaaS, you have to make fewer decisions, but that comes with a cost: you don’t have as much flexibility. You are buying into the platform. That might be great. It’s all just there. It just runs. But I think the costs are not well understood yet.
 
I would probably say I think 2013 is a year of experimentation with PaaS. I would build out a few apps. I would gain some experience. I would look at some pricing models that do some simulation or do some calculations based on apps. I think there is a huge opportunity, but I think there is also a huge set of unknowns that are always a challenge.

Q: So take advantage of a year of experimentation before the forces push you one way or the other and catch you unprepared?

DW: That’s exactly right. Don’t embark on something in a gung-ho way without actually putting your toe in the water and seeing how hot the bath is. That is what I would recommend.

Q: What are some of the implications of all of these different trends on Agile development?

DW: I think that in 2013, the things that will be cool about Agile will continue to be continuous delivery and portfolio management. They will be the hot topics. I think that, in terms of how you organize teams and how you use Agile at that team level, everybody agrees. I think most organizations have daily standards. Most organizations now have backlogs. Most organizations now have burn-downs at the team level, but then how it fits in is still around portfolio management and delivery, so things like continuous delivery and DevOps will continue to be the topics du jour.

Q: It seems that water-scrum-fall is the norm. How do apps VPs find that balance between blind adoption of the Agile manifesto and being so wishy-washy about it that they are not even really doing it?

DW: That is the $64,000 question. We use the analogy to religion. We say Agile is like a religion. Most religions tend to be adopted in a very blind way initially. Over time, pragmatism comes in as people learn that it is the intent of the law and not the law itself that is important. What we find is that organizations are always saying, “Well, that is not very Agile, but it is reality.”
 
I guess there are a couple of things that you should do. Number one is you should employ people who have been through this religious conversion a long time ago and now have a pragmatic approach. Number two, don’t throw out all of the practices that you have held in good regard for decades. But do challenge some of those existing practices, and at least understand why you are doing them. Understand the intent rather than just repeating the law.
 
For more on Agile development, see our November article on Forrester’s look into Agile best practices, and visit hp.com/go/agile.