I was in Las Vegas earlier this year for the Pink Elephant ITSM conference. I took some time out of the conference to meet Dave Gordon, The Practicing IT Project Manager. We discussed IT project management, the challenges for project managers working on Software as a Service projects and the cartoon laws of physics.
This video was filmed on location at The Beat Cafe on Fremont Street, Las Vegas, a very nice coffee shop and the only place I visited where food and drink didn’t come in bucket-sized portions.
Dave Gordon: It’s not that I’m practicing to become perfect. I’ve accepted the fact that perfection is overrated. But as someone who is doing actual IT projects and making a living doing it, I’ve been doing that for over 25 years. I like to identify myself as a practitioner. I’m not a theoretician. I’m not some guy writing textbooks. I’m a guy who basically is out there trying to figure out how to get stuff done. And because I come from a technology background, I’ve seen many waves of technology. I wrote my first program in 1973 on Fortran on a model 29 card punch.
So to me, I’ve had the chance to watch a lot of waves of technology come through. But the one thing that has remained constant in all of that is there’s always a difference between what we think we’re going to do and what we end up doing. Because along the way, we discover an awful lot about what the real problem is. I don’t think there’s a way around that. I think we have to accept that. And I think that’s one of the reasons why the Agile approaches are becoming so popular because they do acknowledge that you can’t plan it a lot and then do it. You can certainly do that in civil engineering because if you’re building a bridge from here to there, here isn’t moving and neither is there. So you’ve got a lot more ability to control the subject matter, the content and the scope.
When you’re talking about building a software system to accomplish some business need that may or may not closely defined that may have a lot of constraints that are external to the organization for compliance or any number of other things. And when you’re talking about technology that seem to refresh every few years that becomes a lot more difficult. So the timescales are necessarily compressed. Things are continuously changing, the parties and interests change so it’s a very different model and I think that that’s really what sets IT project management apart from other forms of project management.
Tell me about the book you are working on…
Dave Gordon: It’s about implementing Software as a Service. Essentially from the point of view of the customer who is trying to replace whatever their legacy system might be with Software as a Service as opposed to a premise-hosted software package and other things that goes along with it.
The basic idea behind the book is that the reality is of the SaaS implementation is it’s less about information technology and far more about re-engineering business processes, really looking at how you represent data in a complex system.
Many years ago, a group of undergraduates with far too much time on their hands created something called the Cartoon Laws of Physics. If you browse the internet, you can find the Cartoon Laws of Physics and one of them is about cats. Cats assume the shape of their containers. I mean if you chop them up into little pieces, they assume the shape of the container whatever it is, a violin case, a jar, whatever and you take them out of the container and they retain that shape. The same thing is true of data. Data also assumes the shape of its container. Whenever you store records in a particular system whatever that system is, you’re basically forcing it to fit in the descriptions that were created for that software application that data model as it were.
So one of the challenges you have whenever you move from some legacy system to a new system is you’ve got to basically reformat that data so that it means something meaningful to the new system. So part of the challenge is going through your data assumptions, your data models and determining what’s actually meaningful in all of that stuff, figuring out how to bring it across to the new system and not do any damage to the meaning of it. And I think that’s one of the key challenges.
Another key challenge is figuring out how you’re going to support your users when in fact there is no in-house technical knowledge so that’s another piece of it. Testing is another piece of it. Security models are another piece of it.
One of the challenge that most organizations have nowadays is they’ve invested a lot of money and some sort of identity management and provision and deep provisioning systems. So how do you integrate an application that’s hosted at some third party data center by some vendor with your security model? So it’s a number of challenges and I thought it would be time for a book like this to pull all those things together.
What are the challenges for project managers on SaaS projects?
Dave Gordon: One of the challenges you have is a very different group of stakeholders. Quite frankly if everything is in-house, it’s a lot easier to identify your stakeholders because you can really very efficiently and effectively say: ‘Okay, who are we talking to now?’
But as soon as you start hosting something outside, you got to integrate that system with all the other systems that the legacy system touches today. In my case, I go around implementing a human capital management and payroll system and they typically have anywhere from 20 to 70 integrations with other systems. Third party vendors, you know people who report taxes withheld on payroll, things of that nature. So add in all of the other internal systems from your accounting system to the aforementioned security management systems, you very quickly end up with a lot of integration work. So it becomes difficult to understand and manage all of those stakeholders because everyone of those sources or sinks of data has some sort of an established set of protocols and some of them you may be doing for the first time and some of them, you may be redoing after having done it 10 years ago and everyone associated within that. So sometimes, it’s about rediscovering the way and you don’t always necessary want to pave over the cow pats. A lot of challenges for that kind of work simply because there are more parties of interest.