Thought Leader Interview
Michael Krigsman is CEO of Asuret and IT Project Failures blogger for ZDNet. As an industry observer he’s been studying IT project failures for a long time. Not long ago he founded Asuret to help companies do a better job and succeed more often in delivering complex IT projects. We caught up with him last week to ask about what can be done to avoid IT project failures. Here are his insights.
Denis Pombriant: In a nutshell, can you encapsulate what your research is about?
Michael Krigsman: Statistics tell us that 30 to 70 percent of IT projects fail in some important way. They’re late, over budget, but most importantly they don’t achieve the intended outcome. These failures happen despite the fact that we have various metrics and methodologies expressly intended to run successful projects. Therefore, we must conclude that something is wrong. The level of waste is very high, we’re talking about billions of dollars every year—not counting opportunity costs.
DP: Have you noticed any typical or common things that companies do when they implement major systems that lead to failure?
MK: We have very sophisticated tools for tracking resources and so forth—but there’s a more fundamental issue most companies don’t address, which is getting a clear picture of expectations that stakeholder groups have of the project. What do they want to get out of the project, and very importantly, as the project goes through its lifecycle we need to find out what ‘s really going on. So for example, very often we see a project where management says “We have this unexpected problem.” But somewhere deep inside the organization,someone knew that warning signs existed. The big challenge is developing a more adaptive, or real time, approach for uncovering this information early on in a project.
DP: This sounds almost like crowd sourcing.
MK: In a broad sense it is, but it’s actually finding the needle in the haystack. It means systematically and consistently getting a handle on people’s perception of the project. What’s the actual, real, genuine state of the project as it’s going through its lifecycle from beginning to end.
DP: So without filters, no one blowing sunshine at you?
MK: Yeah, getting an accurate handle on what’s really going on in a highly granular, accurate, and systematic way. It’s very hard to do that.
DP: Can you name companies or projects where that was done?
MK: I think anytime you have a successful project, then they must have been gathering accurate information as the project progresses so that they could adapt and change and shift.
DP: So are you saying that the current state of the art is that we won’t know if it was a good project until it’s over?
MK: Or until something blows up.
DP: Wow! So, how do we get prescriptive, and diagnose a problem to avoid a project blowing up?
MK: I think it requires a cultural shift thinking about how we construct projects and how we run them. Now, we could easily get into a long discussion about alignment between business and IT, and it would sound like a lot of clichés, but in fact that’s the source of the problem—lack of alignment. That’s the fundamental problem. Another term we hear is talking across each other, talking at cross purposes. So the trick is to get the various stakeholders to be of one mind about what they want and where the project is going.
DP: How do you do that?
MK: That’s what my company, Asuret, does. I don’t know of anybody else who’s developing systematic tools to accomplish this. We’re not dealing with resource issues or quantified categories of budgets, but we measure and examine perceptions. So we need to look inside at the intersection of project mgt, portfolio management, strategic planning, organizational development, change management, and yes, it is a social issue so to speak.
DP: Sounds like a lot of data. How do you collect and collate and analyze all that data?
MK: Asuret builds specialized tools for accomplishing just that.
DP: How’s that going?
MK: It’s going very well. It’s taken the market a long time to recognize the truth of what we’ve been saying for years. But just today (November 9, 2010) Gartner came out with some predictions for 2011, one of which says: [in a section called] Seven Major Projects CIOs Should Consider In The Next 3 Years, “Balancing cost and innovation with risk and governance.” The market finally recognizes that IT-related waste is a serious problem.
DP: Risk and governance?
MK: Yes, to a large extent, although not entirely, these are governance issues. Gartner calls it “business change governance.”
DP: Do you think cloud computing and the subscription economy are inherently better at helping companies avoid implementation pitfalls?
MK: I don’t think they’re inherently better but in practice they are better. Any time you can break your projects down into smaller components and then bundle them together as a portfolio helps you construct a way of managing smaller pieces. We all know that it’s easier to manage smaller pieces than a single large one.
DP: Sort of like eating the elephant one bite at a time?
MK: It’s more tasty that way. I think the cloud tends to encourage a type of agile form of development. The practical impact is making projects smaller, but it’s also about getting the business side, which specifies what it needs and the IT side, which is responsible for actually taking those business goals and translating them into systems, onto the same page.
DP: Or in some cases IT is not involved, it’s business users defining the systems.
MK: Sure, if you’re talking about configuration at that level, but I’m talking about more complex traditional projects that require IT involvement. If a project doesn’t need IT at all then you’ve cut the middleman out entirely. Companies like Workday and Salesforce make that possible to a greater extent than before.
So, it’s the ability to rapidly configure a system, which is typical of the latest SaaS vendors. Basically, point and click means that IT can sit with the business users and go through a prototyping process right on the spot, so the business people can see the impact of changes on the system; that makes everything more concrete for both the business and IT. The business can see the impact of changes and IT can understand what the business is getting at; that’s good for everybody.
DP: Anything else?
MK: There’s another notion that’s related: packaged or productized services, which is going to be the future of many professional services. The notion is that rather than reinventing the wheel with every customer, figure out what you’re trying to do in very narrow terms. Then, break your service spectrum into packages or bundles based on fixed time, fixed scope, and fixed deliverables. SAP is doing a really good job of this by the way. Productized services are very important because the goal is really the same: take a larger project and break it down into smaller and more manageable pieces. It’s going to reduce your risk and uncertainty and create more checkpoints.
DP: Turn nebulous risks into manageable probability?
MK: Yes, that’s a good way to put it.
DP: Avoiding Black Swans?
MK: You’ll still have Black Swans, but fewer of them.
DP: Ok, this may be our last question. Are we getting better, worse or staying the same when it comes to implementation failures.
MK: All things being equal I think we’re basically the same.
DP: After all those decades and all those projects?
MK: Well, look at the statistics.
DP: Is that because we haven’t learned anything or is it because new people basically get into new projects and just make the same mistakes?
MK: I think it’s a couple of things. At the start of a project, everyone is full of hopes and dreams and nobody thinks failure can happen to them. There’s a kind of wishful thinking and denial that occurs at the beginning of most projects. In addition, we just haven’t developed any new tools that drill into the core issues of perception and expectations—and these are perception issues. Success and failure is fundamentally a question of perception. If I’m paying for a project and I think it’s successful then it doesn’t matter how much money we spent, it’s successful. It’s much easier to focus on resource loading and scheduling than it is to focus on, and measure, perceptions.
DP: Hopefully you can change that perception. Many thanks, Michael.
MK: Any time.
###
