Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage Interviews Mike Griffiths on DSDM, Agile and the PMI

Mike Griffiths on DSDM, Agile and the PMI


1. Hi, I’m Craig Smith, I’m an Agile editor at InfoQ and we’re here at Agile 2012 in Dallas in Texas, it’s my great pleasure to be here with Mike Griffiths. How are you doing, Mike?

Good thanks, how are you, Craig?

Craig: Good, thank you. So, for those who don’t know you, you’re an independent project manager, you’re a trainer and consultant, so, tell everybody what your basic background is.

My involvement in Agile worked out quite fortuitously, I just happened to be working for a company called Data Sciences in England, back in 1994, and Data Sciences were one of the companies involved in the creation of DSDM. So, DSDM the Agile method, was a consortium based method and I was working there as a developer at the time and later as a team lead, and so I was just in the right place at the right time and got involved in the creation of DSDM, I enjoyed the process and stayed with that method as I worked for IBM and we enjoyed using Agile approaches towards solving problems. My previous background before that was in more rigorous defined military project development, which was very much high rigor high process and so to be able to balance that or bring the pendulum a little bit more towards user involvement was really good. Yes, I got started in 1994, I’ve been using Agile for the last 18 years exclusively, so certainly been immersed for a long time, it feels now, in the whole Agile community which has been good for me.


2. That’s great. So, I would just like to pick up on the military thing for just a minute, having come from that background and looking back on that, is there anything that we can learn or still need to learn from perhaps the way the military worked?

Some things are very different, some things are quite similar. Things that are very different; we use formal specification languages like Z and OBJ which were really rigorous in gathering our requirements and I think we’ve come a long way on from there with the user story format, but at the same time we still have a lot of the rigor, we were looking for traceability from tests to each requirement and I think now in our tools and our unit testing where we have the test per requirement or story and looking for 100% code coverage in test driven development, in a lot of ways they’re different but in a lot of ways it’s similar goals, I think you see that in industry anyway, the same problems come up time and time again and the way we solve them perhaps varies.


3. You mentioned DSDM, probably a lot of people who will be watching this, even people at this conference wouldn’t be aware that a lot of the original Agile methods were invented in the 90s, and DSDM was a pretty big part of that. What’s different between DSDM as it was and where the Agile community is now? What are the big differences?

Well, the big difference was that in the early days DSDM was like a paid membership model, and I think that really hurt the methodology a lot, so as Scrum and other methods were being popularized and made freely available, you had to pay and become a member of this consortium to get access to the methods and I think the big learning message from the whole thing is to make it available and try and gain market share and that’s the way to grow and be successful, not to have a closed model that you have to pay to get access to. So, many similarities in terms of prioritization, maximizing the amount of work not done, so in other words looking for the simple solution inside a big complex problem, and Arie van Bennekum who represented DSDM at the original 2001 Snowbird meeting proposed DSDM’s nine principles as a starting point for the Agile Manifesto, so obviously a lot of smart people there had a lot of other great ideas too, but you are right, it did actually have a strong influence on how Agile really got started.

Craig: And some of the ideas are, having been at that Snowbird meeting and having DSDM representation there, some of those ideas are now baked into what we doi now?

For sure, yes.

Craig: So, you mentioned the open and close thing, I wonder whether back in the 90’s, we didn’t really have the internet, at least not as available as it is now and in those days, the things that you were competing with, things like RUP and those type of competing approaches, they are all in that same model.

Yes, that’s right. I think it’s hindsight that shows that perhaps the methods that did really well were those that were open, and you are right, we were competing with RUP and we were competing with METHOD/1, and SSADM, and they were all large paid for models because you got a lot of collateral for your membership fees, yes.


4. I’m curious just on the DSDM topic, how was it that a bunch of you guys got together to actually create DSDM in the first place, what was the interest that you saw?

Well, actually it was pretty similar to what happened with Snowbird. There was a number of companies in the UK who had been using James Martin’s RAD approaches, some elements of that worked, like the feedback loop shortening of prototyping and then getting that feedback, that was working really well, however building prototypes in throw away tools that were more front ends and things and then having to go re-implement in the final language, it wasn’t working so well, it gave the impression the solution was near complete but then you had to explain “no, this is just smoke and mirrors, this is just a UI and we actually have to go build this, maturely”. So, people were finding good things about RAD and prototyping and yet at the same time they were stuck with challenges and they had to solve those, and then we had some other good work going on like Weir and Mumford’s participative design approach, getting the users involved in the creation and design of things. So it was really the confluence of companies trying to overcome the limitations of RAD along with some new ideas really just prompted this collection of companies to get together and collaborate, much like how the Agile Manifesto came to be.


5. Cool. So, DSDM is still in use, around the place and I believe there is an open version of that out there now. What’s the interest for people still using that, is it a historical thing or is there actually still organizations where that is actually the best approach for them?

Sure. I don’t think it’s just the kind of a nostalgic “I like to be old school and use DSDM”. It really has some elements in there that suit some organizations well, so, whereas many of the methodologies like Scrum and XP and others do really well from “here’s our candidate requirements list or feature list, how do we translate that into a high quality tested product ready to deploy?”, DSDM lifecycle coverage is much broader, so it has elements in there for project feasibility and other roles like architecture. So companies that are perhaps looking for a methodology with a broader coverage might look for DSDM for guidance and things like feasibility. And so that’s why some companies use it, it’s a better fit for their organizational structure.

Craig: I assume following something like Alistair Cockburn’s Oath of Non-Allegiance, some of those things where you can just pull out and borrow from the process.

Absolutely. And I think that’s what most people do, I am particularly not a very religious zealot about my Agile methodology practices, if I find something totally cool in Feature Driven Development like the parking lot diagrams we were talking about, great, why not incorporate those? So, it’s the same way, if you are running a Scrum project and you are looking for a feasibility artifact or a subset of tools to apply, take a look at DSDM, maybe it’s just a ticket view or maybe you can look at it and say yes, we can do better and adapt it and create something for you.

Craig: So, in more recent years you’ve became more involved with the project management community which is probably where most of our watchers know you for these days. The first thing is that you are both a Prince2 and a PMI Practitioner.

Yes, that’s right. So, working for IBM in the UK I got my Prince2 certification and I worked with DSDM and the Prince2 group creating some white papers, how do we use DSDM or Agile in a Prince2 environment and that was useful bringing those two camps together, but then 12 years ago I moved to Calgary in Canada and trying to get a job as a project manager there with my Prince2 certificate and at that time in North America notmany people had heard about Prince2,so I went and got my PMP certification. And that wasn’t a big transfer because they cover very similar domains, so I got my PMP and then got engaged with the PMI community. So I think it was 2004 I got a paper accepted at their global congress which was presented on using Agile alongside the PMBOK guide. And for a number of years I’d suggested titles such as Agile as a replacement for the PMBOK or The PMBOK is wrong, you need to be doing Agile, I don’t know why I was so slow in learning, but those papers never got accepted and it wasn’t until 2004 when I wised up and did this Agile alongside the PMBOK that I got accepted and presented that and then they asked me to create a two day training class as part of their SeminarsWorld training program, so for the last six or seven years I’ve been teaching that Agile Project Management for the PMI. So, that’s been a lot of fun.


6. For most people that maybe are looking into going into project management, maybe not even in an Agile world but obviously that’s the people who will be watching this, why would they choose PMI over Prince2, I mean certainly from what I’ve seen, countries tend to focus on other ones, me being from Australia we tend to find governments go one way and companies go the other, but what would be the main reason for choosing one or the other?

I think it’s probably organizational acceptance, so if your company is following Prince2 guidance or you are doing work for clients that are Prince2 in their project management, then you probably go that way, whereas if you are a more of a PMI shop then you go that way. Interestingly, I think there is merits in looking at both of the methods to see what portions you can choose.


7. You mentioned that back in 2004 you were already starting to advocate this whole getting Agile, or its other paths, maybe Scrum at the time, working in correlation with the PMI, how was that accepted because it’s been I assume a long path to get some of the big organizations onside?

Yes, it has. And it feels a bit weird as well because for some people the PMI is what we have been rebelling against as being in the Agile community, so why are you working with the enemy or the PMI, this big evil empire? The real answer I think is that there’s half a million PMP project managers and they didn’t have a lot of guidance from the PMI in terms of “well, my team wants to do Agile, I still got my job as a project manager, how do I reconcile their requests with my obligations to be effective or successful project manager?” How do we bridge those two things together, how do we bring the communities together? So, that’s what motivated me, you could either just ignore the PMI, hope that they’d go away, ridicule the old school thinking about projects or work with them to try and introduce some of these new techniques. And so it wasn’t just me, there was many people that worked in the creation of some of the more modern offerings that the PMI now has, so yes, we decided to see if we could go change them as opposed to work away from them or just ignore them.


8. [...] What’s the thought people like you or practitioners in the project management community about that, how would a Scrum team, someone who is doing pure Scrum, then take on project management?

Craig's full question: And it always tends to be a good result when you do that. So, just before I go on to where that’s been going, one thing I have always been curious about is that if you look at Scrum in particular there is no mention of the project manager at all, they tend to think that the Scrum Master and the Product Owner are the only roles you need. What’s the thought people like you or practitioners in the project management community about that, how would a Scrum team, someone who is doing pure Scrum, then take on project management?

Yes, and you’re right. Originally, at least in Scrum you just had the Scrum Master that maybe like a technical lead, and I have worked with projects where we have technical leads that perform that role and that’s great and then as you scale those teams, typically you need somebody to act as a coordinator between them or as a bit of a conduit to business and different stakeholders and that maybe where the project manager role fits more in. So it’s, yes, letting the technical team do their work but then assisting them to remove impediments that aren’t in their circle of control and also spreading the message about what they are doing to a wider set of stakeholders.


9. Would you suggest then that people who are reading, say the works coming out of Scrum, even the latest updates where there is still no real mention of a project manager, that the role still exists, it’s just that it’s outside the team view that Scrum typically has?

Yes, you can view it like that or you could have a project manager as Scrum Master, I’ve seen it work both ways, and if you don’t have a need for one, great, why invent one, but typically most organizations, large enterprises in particular do have a need for that kind of a role.


10. Is there any danger in a Scrum Master being a project manager or vice versa?

No, Scrum Masters are looking for career paths as much as anybody else, I think just as long as they are fulfilling the roles of a project manager and the stakeholders are happy, yes, all for it.

Craig: So, it’s more about the skill rather than the title.

Yes, sure.

Craig: Awesome. So, in recent years you’ve been a large contributor to the PMBOK and well known for publicising a lot of that work going out there, most recently version 5 has just been published I believe or is in course of being published and there is now a lot more Agile acknowledgement.

Not really a lot more, but certainly more. I’d like to see even greater amount of Agile acknowledgement, but yes, the PMBOK version 5 it is out, like you say, for review at the moment is going to be published in January, so we have got Agile actually in the dictionary if you like and defined in the glossary of terms and a few chapters mention it and how Agile methods tackle portions of the project. But, it’s still small and I think that is probably rightly so, because that PMBOK guide is industry agnostic so it’s not only for software projects but it’s also for civil engineering projects, medical projects and the application of Agile is growing outside of the software space, but it’s really focused in the knowledge worker domain, and the PMBOK is for knowledge workers and also more traditional widget manufacturers and everything else. So, you have got to understand the limitation of Agile spread into the PMBOK guide. There is actually a software extension to the PMBOK guide that we are working on as well right now, and that has heavy, an extensive Agile coverage. So, if you are looking for where’s the Agile in the PMBOK guide I would direct people to the software extension to the PMBOK guide. So, we’ve got a previous model, there is a construction extension to the PMBOK guide that says “hey, all this PMBOK stuff is good, but if you are doing a construction project then you might want to consider planning like this and procurement engagement like this”. So, following that model we’re saying “hey, here is a software extension to the PMBOK guide that says you might want to plan and estimate and track your project in more of a software-oriented way”. And that’s where the Agile content really comes in.

Craig: Even for people who just pick it up and read it, I guess this is what you’ve been talking about since the early days, all the stuff that’s in there it’s still applicable to whatever method you are using. So, people who read it as a methodology in its own right, are probably not getting the intention in the first place.

Exactly, yes.

Craig: The other thing that I’ve seen coming up recently is the notion of this Agile Certified Practitioner. Tell us a little bit more about that.

Yes, so the PMI said they were going to get into the Agile space and provide not a body of knowledge but some skill areas and some education for their members, and they have done that and we have the Agile Community of Practice and offerings its built out. But then they also wanted to get into the certification space and that, too, was kind of a double edge sword for me, and Agile certification from a PMI, that’s just an oxymoron. And so I looked at that and I thought well, it seems like it’s going to go ahead, so we can either be on the inside trying to steer it in the right direction or on the outside hoping it will go in the right direction or quietly die a quiet death. And so when I heard about some of the other people interested in that, people like Alistair Cockburn and Michele Sliger who wrote the bridging book for project managers, it was yes, I really want to be a part of this and help steer it in the right direction. So, the actual ACP certification of Agile Certified Practitioner was not a PMI sourced or created project, they totally gave us as the subject matter experts, free rein.

So, it doesn’t refer to the PMBOK guide at all, it’s not a test about anything in the PMBOK and there are 11 reference books, works by Cohn, Jim Highsmith, that are real fundamental pieces in the industry, and that’s what it tests people’s knowledge against. So, it was good that we were given that free rein and they recognize that, and I think it’s a good certification. So, how it differs from things like the Certified Scrum Master is that instead of being focused purely on Scrum, it’s a broader certification. So, there’s Agile in there, but there is also Lean and Kanban, so it’s not just testing your knowledge on say XP, you’re expected to know XP and Scrum and Lean and Kanban, and be exposed to a wider set of methods and we also acknowledge that probably most projects have some degree of geographical distribution so we are not thinking very small team co-located, it’s most projects most of the time have one or more person working remotely, so how do you deal with that? It’s a broader set.


11. And people don’t have to have been a PMI certified in order to get that?

No, absolutely not. You can come straight in and get ACP and not touch any of the other PMI certifications, if you just work in the Agile domain.


12. Is there a suggestion that you should do it one way or the other?

No, and I was chatting to them at the PMI booth here at the conference and there’s about 1,200 people in there who got their ACP and certainly in the early days two-thirds didn’t have their PMP, so it was sort of Agile only. I don’t know what that break down is now, but in the early days it was that.


13. Is there some recognition that maybe people have gone through and done that that then they may come back and do some sort of project management certification afterwards?

I don’t know actually if that’s a hope, I guess as a certifying body they’d like to think they’ll get some of that cross pollination.


14. Good. So, one of the things you’ve done in conjunction with this is you’ve actually published a book to help people through that journey called the PMI- ACP Exam Prep, correct? Looks like a good book to actually help you with that, so how did that come about?

Sure. Well, thanks for that. It came about, I was approached by RMC, so that was Rita Mulcahy’s original company, I read her book when I moved to Canada and was sitting for the PMP certification and I really liked the format of that, so when I was approached to help them create the ACP guide, it was great but it was also a little bit daunting, because I knew the standard of that preparation set of materials. But they had a great team and I worked with them over about a nine month period to create that guide and so far all the feedback we’ve had has been really positive, people like the book, they find the case studies and texts kind of entertaining, which is nice, because studying for an exam can be a daunting prospect, so if you can make it more human and accessible then that always goes down well, because you don’t need on top of the pressure of sitting an exam some big scary book, and I know Rita’s book was great because the PMBOK guide was kind of cold to read and so she brought a human element to it and that’s what we tried to do as well in that ACP Prep Guide, to have lots of examples, lots of test questions and exercises.


15. And unlike maybe some certifications that people have done before, I guess the nice thing about that particular book is that it goes alongside all the other texts that are in that space, right?

Yes. What I’ve tried to do is bring those 11 textbooks together, the content that was used in the exam, content outline, and just sort of introduce them with a common voice and then speak to some of their differences because the roles and the techniques, sprint vs iteration, reading those 11 books you can get different interpretations of similar concepts and so we were trying to use this common voice to explain the concepts.

Craig: Great. So the other thing you have also done is help start the PMI Agile Community of Practice. Tell us a little bit about that.

Yes, that’s really exciting, that’s got over 13,000 members now, it’s nearly the largest PMI community of practice and my role was actually very small. I presented at one of these conferences where we were talking about bringing Agile to the PMI and I was approached afterwards by Jesse Fewell and he said “wow, this is great, we should do this, we should go do it”, and at that time I had been struggling for years to get my Agile and PMI stuff accepted and I basically said to Jesse “yes, good luck with that, tell me how it goes, I’d like to help you but I don’t have capacity for that at the moment”. And good on Jesse for driving it forward, building the community. I was involved with that, but it was really Jesse that did it and went through all the red tape of creating the community of practice, because at the time the PMI used to have special interest group SIGs and they were just changing the model over to communities of practice, so not only did we have to figure out what to do, but then we were the first community of practice to push through that process and he moved with it and built a community and yes, it’s a really large active vibrant group now, which is great.


16. What would one of our people watching this get if they joined the Agile Community of Practice?

Sure. Well, the good thing is it’s free, so if you’re a PMI member it’s free, it’s a great resource. So, as a project manager perhaps you are new to Agile, and you say “what does an Agile plan look like, how are people doing risk management or procurement”, and you could go there and you could look at existing case studies and artifacts, but the main thing is the community itself. So, you could post a question and there are all sorts of experts and people in there posting answers, so you can get your questions answered, you would find similar like-minded people solving the same kinds of problems that you have. So, I think that is the greatest thing, if you are a PMI member it’s free, you can go in there and it might be able to save you a bunch of time on your project.


17. [...] So, what’s got you interested in metrics, is it that project management side where you had to show the output of projects?

Craig's full question: That’s great. So, one of the ways that I have come to know about you is through your blog, which is, which has a great amount of useful content and we were talking about parking lot diagrams before as I mentioned as a coach people often ask me how do I track progress and I start mentioning parking lots and they go “how do I find out more about that” and about the only place I can find anywhere on the internet is your blog, so that’s probably kind of what got me started there. You’ve actually really got some really good stuff on there, I guess the things that stood out to me is that you do a lot of work on metrics, amongst other things, you do talk about project management stuff there as well. You’ve got some really interesting things in there, stuff I haven’t seen before, things like you’ve got the Agile Smart Metrics which is pretty good, and smart metrics is something that you spoke about before. So, what’s got you interested in metrics, is it that project management side where you had to show the output of projects?

Yes, I think it is and also what really got me interested is how many dumb things projects track. Working for large companies over the years and moving around, you get to see a lot of metrics and then you get to see the potentially damaging effects of having those metrics so tracking things like lines of code written or hours worked or budget consumed, just leads to more hours worked, not necessarily a better product, right? Then people get burnt out, lines of code written, how are we going to reward simplification and refactoring if it appears like the work we’ve done has now suddenly got smaller and so just being exposed to the unintended side effects of many of these metrics just peaked my interest and got me investigating it a little bit more, and then I read some great books like Don Reinertsen’s Managing the Design Factory and Principles of Flow, and it really just illustrated the big mismatch between generally accepted project management metrics and what we should really be measuring. So I think it’s just a personal interest area.


18. One of the most recent posts you had which I was reading just recently is talking about taking risk management and applying sort of collaborative games to that, you’ve got a multi-part post on your blog about that. So, can you tell our listeners a little more about that?

Yes, so I am really excited about that, I’ve been putting risk reduction items in my backlogs for a number of years, but I have been doing that as a Scrum Master and project manager, it wasn’t until more recently I got the whole team engaged in risk identification, finding a possible risk for a project and also opportunity, the plus side of projects, too, and trying to make opportunities happen and leverage, I really acknowledge the benefits of this whole collaborative approach, and it makes a lot of sense. We know not to let the project managers estimate projects because they don’t have the technical insight so we use games like planning poker to tap into their collective knowledge. And the same goes for many of the risks on a project too, especially technical risks, project managers don’t have the best insight into where those risks may be so why not engage a team in figuring out where they and also opportunities maybe there are some quick wins that we can get here.

So, instead of having just one person like a project manager, you’ve got a team of eight or ten people looking out for risks, trying to avoid risks and also looking for opportunities as well. I’ve just found that it’s been a huge enabler for our projects, they’ve just gone more smoothly, we haven’t had as many issues, because these risks haven’t happened and I think it’s really just getting the team up to this level of awareness, making sure everyone is aware of the risks on a project and they do small subtle day to day things to make those risks go away. And they never come to me, and I don’t think it’s just been pure luck but in the last year where we’ve been using this approach in companies, more of the opportunities have been realized, fewer of the risks have actually occurred, I like to put it down to the sort of a collaborative approach of raising the awareness of these things throughout the team and using that collective knowledge I guess to steer the projects towards success. I’m really excited about that. If you look at the levels of magnitude, whether it’s process or people, if you look at any of the COCOMO software estimation metrics, they say there is an order of magnitude difference between having average level people on your project and best in category, whereas the process multipliers are a factor of ten smaller.

So, going from average Scrum to the very best Scrum will get you an increase of x, but having a good team to a great team will get you a 10x improvement. And so for me this whole collaborative risk management is investing in the people side, getting the people a little bit more aware of some of the issues and then tapping into their experience and ability to go influence stuff. So, I am excited about it because I have seen really good results and in about two hours here at the conference I will be giving my little spiel on this collaborative games for risk management.


19. Awesome. So, I was just going to move on to that, you’ve got two talks here at the conference here in Dallas, one of them being “Risk Management is too important for Project Managers”, is it exactly about what we talked about?

It is. We’re walking through the games that we introduce to teams for risk management, so we have games around getting the awareness of the importance of risk and opportunity management and then walk through all the steps of finding risks, classifying them, quantifying them, deciding what to do about them, making sure they get into that backlog, so it’s like a tradeoff conversation we have with the Product Owner because they own that insertion into the backlog, so how do we convince them it’s important to get them in, and then ways of tracking and reporting on risks as we go.


21. You’re suggesting we should not call our people resources anymore, which we should’ve done anyway, and call them cows now?

We should call them cows now? It’s no worse than pigs or chickens, I guess. Really, it was a poor way of introducing leadership topics and I had to have a bit of a tie into the whole Dallas theme, so it was the Dallas cowboys, the whole resort here is very much cowboy themed, so ok, how can we get to speak about leadership topics, and so that was the key into that, and the main thing is people are the biggest part of our projects and most likely success source for us, so if we can find ways to motivate them, to support them, provide what they need, this is, like I was saying before, really where I think the greatest leverage is and me as a project manager I found my greatest benefits occurred when I started listening to my team, finding out what they needed, finding ways to support them and grow them. And so, I have been deeply ingrained in Agile and it’s about processes and tools and that’s all cool and as an engineer I like to tweak and twiddle but it comes down to the people. So, if we can find a way to get them fired up, excited, keep them happy, make them want to work on our projects, that’s really where for me the benefits are. So, my talk a couple of days ago was on leadership which is really how Agile teams should be looked after, rather than managed and especially micromanaged, we want to inspire and lead our teams.


22. One last question I want to ask you before we leave is having been at the conference, spoken to lots of people, talking about things we have, what’s next for us as an Agile community, what are the next big things that as community we need to be tackling?

I think we’ve got the movement beyond software, so we recognize that software is a knowledge worker community so we collaborate on problems that organizations typically haven’t addressed before, we work with partial information, we try and solve problems and we manipulate information as opposed to concrete and steel or whatever it is. But that’s not unique, all other forms of engineers do that, doctors and teachers and other people manipulate information, have incomplete information, so software engineers are in this knowledge worker space, we have some good sets and tools and I think for Agile as a community will be to take it outside of software but also look at these other techniques, how do doctors share information, what can we learn from them and bring it back into our community, because they have a lot of unstructured information and good ways of recording what works and what doesn’t work. So I am hoping personally that we will expand beyond software more so, I know it’s been happening anyway, but not only will we spread our message, but we also learn and bring some new ideas in. And the conference has been great for me because I have been exposed to some new ideas and things that are coming from outside of software, so I am really excited about the future, see how it will blossom.


23. That’s great. So, if people want to know more about you or find out more of your work, where can they go?

Sure, is my website; you can find references to how to get a hold of me there.<

Craig: Well, Mike, thank you very much for your time, it’s been a pleasure.

You are very welcome. Thanks.

Mar 22, 2013