[Reposted from InfoQ]
The InfoQ editor Chris Goldsbury reports a Friday afternoon conversation with some noted agile practitioners covering the current state and future trends of the methodological universe and how these are playing out in the real world.
InfoQ: The Standish Group for 2011 is reporting that project success (on time, on budget, with functionality intended) improved by 5%. From your vantage point is this a result of increased agile adoption in the market place? Are there other factors at play? Are we getting better at delivering software development projects consistently?
Brad: My view is that clients are being more careful about how they approach projects in general and that’s probably what’s leading to the improvement. I wouldn’t correlate agile to this improvement. It’s hard to differentiate between those who are really doing agile and those who are doing faux agile. So many times we come into a client and find they’ve only partially adopted some practices, and this watered down agility creates the mistaken belief that they’re successful at agile implementation.
Peter: I agree with Brad. Project metrics aren’t getting better because of agile. If you look at the need in the market, how it’s trending, as I do, what you see is a lot of shops doing agile poorly or in a lightweight way. People aren’t transforming their culture; they’re just going through the motions and calling it agile. Standish’s report may give companies a reason to believe agile is better at delivering projects, but that’s not what we are seeing.
Brad: Standish report metrics may be flawed. Mature agile teams are trying to optimize value. Not necessarily stick to the rigid Standish metrics of scope, schedule and budget. If we’re late, but we delivered successful functionality then is that really a failure? If we only deliver 40% of scope but that’s where the major value is…isn’t that successful? Standish doesn’t measure this.
Peter: We need a new metric for measuring software development project success. It’s not just about scope, time, budget, risks, and issues. It’s about harnessing overall value. How does Standish measure that?
David: My major issue with Standish is it is flawed from the get-go. They assume we know the solution from the start. We rarely know this. We can’t use a cookie cutter approach when doing software development and Standish assumes that our initial plan won’t adjust as we move through project delivery.
InfoQ: Is there too much hype around agile / lean or other methodologies? How does this affect your business?
Peter: There’s a lot of hype around agile, lean, and kanban. Hype is good though. The more conversations we’re having around these topics the more it grows. In some ways the hype has led to unrealistic expectations. I get a lot of requisitions for an agile coach at $40.00 an hour, but you’re not going to find a transformation agent for that bill rate and that’s really what you’re buying when you get an agile coach.
Brad: I’m not going to disagree with Peter. Hype is good. But it’s also problematic. I used to be able to walk into an executive office and present agile unbiased. Now, we have to spend time educating them on what agile really is and how it works…..correcting the misconceptions. A startling reality today is the guy who took a CSM (certified scrum master) course 6 months ago is now an agile coach. This destroys the agile value proposition by diluting the industry with inexperienced coaches.
David: I come at it from the start-up perspective. The hype is starting to seep in here too. There’s a lot of writing about agile, lean startup, but not much doing. They’re excited about it, but start-ups need help doing it and need someone to guide them through the hype and into value. Agile, Lean-Startup are worthwhile and it’s a scientific way of building companies. The hype will settle down and people will start finding value and adapting that to the enterprise. It’s here to stay, but there’s definitely a lot of marketing going on right now.
InfoQ: Which methodology works most often in your experience?
David: I tend to mix up the methodologies and use what makes sense. If I had to pick one: XP.
Brad: We’re on record as being solution engineers first, methodologists second. We’re focused on clarity of outcome. Where client constraints don’t stop us from embracing and scrum/xp in combination we’ll do it. But, usually we focus on scrum. If you give me a group of expert developers I could care less how you organize the work. They’ll get it done no matter what. At GearStream we tend to stick with scrum, some xp, but with a lot of modification to suit client needs.
Peter: XP is a foundation for what I do. Scrum is laid out to be very easily understood and has 3 distinct roles, but there’s lots of controversy around the product owner role. I use a combination of methodologies. It does depend on the client and their level of agile maturity. You really have to survey their environment and do what’s best for them. I tend to seek out what’s best for my client and tailor it to them. Great agile coaches are great at leveraging all sorts of facets. Let me just add this: I would love everyone to call this the “way we do work” and drop the agile, kanban and other methodology monikers.
InfoQ: Do you believe methodologies are melting away and we’re focusing on techniques? Where’s the future?
Brad: The name game won’t go away, and we’ll get more of these methodologies in the future. This is good, and we’ll see improvements and new ways of doing software development.
David: We’re headed to continuous delivery. We’ve only scratched the surface on this. You can’t afford to be slow. You need to deliver quicker. I can’t tell you how this will play out in terms of tools, but the trend is clear in start-ups.
Peter: I agree with Brad and David. We have to start doing more deployment, faster. We have to stay ahead of the curve. In the consumer space…people are expecting this kind of quick turnaround.
Brad: I want to tag on to that theme a bit. I agree continuous delivery is where it’s going. But building fast is only good if you’re building the right thing. Stronger product management services and human centered design are the next big stage. Most companies aren’t good at product value management and could use help in this area.
InfoQ: How has the social media start-up explosion helped/hindered your business? Is the interest in agile a result of this explosion or is it a separate organic phenomenon?
Brad: Social media is less about agile and more about finding talent and clients. We stay more connected to our clients using this medium.
Peter: Social media for me, as an independent coach, is essential. I’m in a bit of a unique position with Agile Scout. I get substantial business by utilizing these tools.
David: I’m a big fan of Peter’s social media strategy. I really think it helps sustain relationships over time. I hear from people in a conference who’ve used something I wrote about on my blog. It really helps to hear someone used one of my ideas and gets me thinking about how it could be modified to suit other industries and situations.
InfoQ: What advice would you give a business that is trying to improve its software development process and delivery?
Brad: Start with the outcome you have in mind first. The objective isn’t really about software development. We need to quit talking about software development as if it’s a separate, remote piece of the company. It’s an embedded part of the operation. Make sure the leadership has this in their vision. Two points:
- Clarify how improving your software development process will benefit your business strategy.
- Engage business leaders for being accountable for this in the same way they are accountable for any other part of the business. It’s not separate.
Peter: Kaizen on self. If you don’t have a consultancy or independent coach working with your business to improve it then the onus is really on you to improve. Continually grow your craft. Learn from those people you follow and stand on their shoulders. As coaches, consultants, we are problem solvers and that’s where the value is to any client.
David: I would ask any client: What do they expect to gain from agile? What is your goal? I don’t want to talk about agile. I want to talk about the perceived benefit, what’s the pain point they are trying to solve? Start there and then we can figure out what makes sense for your start-up.
About the Interviewees
David J. Bland – With over 10 years of experience in the technology industry, David has enjoyed success using agile software development at several companies near Washington DC. He joined his first startup in 1999, and helped scale it to a 13 million dollar acquisition in 2006. Since then David has helped teams deliver early and often for products ranging from eCommerce to Counter-Terrorism. David strives to find a balance between agile theory and agile reality, while expanding into complimentary process areas such as leanstartup, business model generation and customer development.
Brad Murphy – As CEO of Gear Stream, Brad is responsible for overseeing the company’s strategic direction, planning, and growth. Brad’s vision for Gear Stream is keenly focused on transforming how companies leverage Lean, Agile and Enterprise 2.0 collaboration. Brad’s vision includes helping organizations transition from project driven governance to Lean, continuous flow, product line execution. While others in the Agile community focus narrowly on core development, Brad’s vision includes a framework for leveraging Agile development teams in a synchronized, cross functional stakeholder workflow including Product Management, Risk & Compliance (PMO), Outsource Partners, Finance and Release / Fulfillment operations.
Brads 27 year career includes two prior start-up successes including a software ISV, and a leading eCommerce Outsourcing and Advisory firm called digitaESP, which was acquired by Valtech, S.A. in 2002. More recently Brad served as the CEO of Valtech U.S. where he helped lead and grow one of the industry’s most successful Agile Advisory and Consulting firms.
Peter Saddington – Independent agile coach and owner of Agile Scout. Peter Saddington is an Independent Enterprise Agile Transformation Coach with Thinqube – Lean Incubator, Inc., a private consultancy. He has been in Agile software development for over 14 years. Previously a software developer in the mid 90’s, he has since helped the Department of Defense (DoD), United States Air Force, Centers for Disease Control (CDC), Johnson and Johnson, Primedia, Consumer Source, Blue Cross Blue Shield, T-Mobile, and various other private and Fortune 500 companies reach agility within their development teams and enterprises. An avid writer, he is the author of the book Scrum Pocket Guide: A Quick Start Guide to Agile Software Development (www.scrumpocketguide.com), and the Executive Editor of the popular AgileScout (www.agilescout.com) news blog that reaches thousands of viewers per day all over the world.