With Agile-Lean Ireland 2019 just around the corner, we had a chat with one of our keynote speakers, Dan North, to find out his thoughts on topics ranging from the future of BDD, the future of Agile, scaling, and testing challenges…
ALI: How did the creation of BDD come about?
Dan: It started in 2003 as a way of coaching TDD. I found people understood the iterative, example-guided nature of TDD more easily if I avoided the word “test” altogether, and instead used words like “example” or “specification”. Early in 2004 Chris Matts and I extended the example-based language from code-level TDD to feature-level, which meant we could use the same approach at both levels.
Contrary to popular belief TDD does this too. People think TDD is just at the code level, and you need something else, like ATDD or Specification by Example, at the feature level, but the way Kent Beck describes TDD includes these multiple levels.
ALI: What is the biggest mistake you commonly see teams marking when they try to adopt a BDD approach?
Dan: BDD is a whole-team method for delivering software by using examples to guide development. I gave a talk a while ago about ways this can go wrong. Many organisations are set up for work to remain in silos, even in supposedly cross-functional teams. Instead of collaborating on work, each sprint or iteration is its own mini-waterfall.
In this environment, rather than the scenarios being a tool for collaboration, they become a passive-aggressive means of passing work and blame between team members. The magic isn’t in specifying behaviour as Given-When-Then triples, it is in specifying behaviour as a team.
ALI: What are the biggest testing challenges that you commonly see with the teams you work with?
Dan: The biggest challenge I see is that teams don’t understand the purpose or value of testing, so it gets de-prioritised or even dropped altogether. This is even true among many of the testers I speak with. I teach a class called Testing Faster which I open by asking the attendees what they think the purpose of testing is. The answers are varied and surprising, to say the least.
Once you have a shared understanding of the goal of testing, it is easier to have sensible, balanced discussions about how much and what kinds of testing are appropriate for a given context.
ALI: There’s lots of talk now around scaled Agile frameworks such as SAFe, Nexus, LESS, etc. with mixed results. How do you approach the challenge of scaling this way of working?
Dan: I made up my own tongue-in-cheek acronym as a response to this: SWaRMing: Scaling Without a Religious Methodology. In general I think they are all solving the wrong problem. They arrange everyone into Scrum-style feature teams and then have different opinions on how to slice up the work into these teams, manage dependencies between the teams, and gather it all together afterwards. Instead I go back to the lean principle of “moving the people to the work”.
I start with understanding the demand, then get the team to self-organise around the work, in a quarterly planning exercise. I have introduced this approach and seen it used in programme teams of 12 to 200 people. Because the people choose their own work there is a higher sense of ownership, accountability and purpose in the teams after the quarterly planning session. We then use OKRs to define the goals for the quarter, and Rolling Wave Planning to track progress. This means talking about what there is left to do and what options we have to do it, rather than ticking off a list of features.
I could go on at length about this, but I’ll stop here!
ALI: What are the biggest challenges you commonly see with organisations in moving to a more agile way of working?
Dan: They are many and varied! Sometimes it is that the senior management don’t know why they want the change, they just know they need to Go Agile, so they have no starting baseline, no way of measuring progress, and no success criteria. Sometimes it is treated like a traditional project with start and end dates, a budget, a programme manager, and the assumption that at the end we will be Transformed and Agile. Mostly, as Eliyahu Goldratt says, it is that people find it difficult to change their existing mindset.
To change ways of working you need to change the system of work that led to the current ways of working. Many people, especially middle and senior managers, are threatened by this, so they carefully resist your best efforts to improve things. Eventually the people championing the new ways of working give up or move on and the status quo wins.
You have to assume the current system of work is resilient, and is optimal for its current constraints. So to change behaviour you need to change the constraints. For instance most organisations are set up to optimise for keeping people busy and locally optimising project budgets. To get different behaviour you need to start optimising for things like lead time, throughput and flow efficiency.
ALI: In one of your past talks, you described a lot of the frameworks and methods in Agile and Lean as “Lean Operations” – what is your definition of this?
Dan: People tend to think of Lean in terms of manufacturing or production, but lean principles apply throughout the organisation, into areas like product design as Lean Product Development, or finance as Throughput Accounting. Lean Operations refers to the practice of applying lean principles to the whole value chain.
Some methods are about applying these principles, most seem to be about implementing practices based on them, that worked in one context, perhaps long ago, as though the practices themselves are an agile talisman. As long as you have Retrospectives, a Backlog, Stand-ups, Story Points, or whatever your particular agile denomination mandates, you will be Agile.
ALI: Where do you see Agile being in 5 years time?
Dan: Five years passes surprisingly quickly. I suspect in five years another tranche of 2-3 year Agile Transformations™ will have come and gone, with the obligatory transfer of funds from client to consulting firm, with the usual certification theatre, and with no noticeable impact whatsoever.
In the meantime number of organisations will have quietly and determinedly been transforming without a religious methodology, looking at where they want to go and being realistic about how challenging the journey will be.
People will still be arguing over “agile” versus “Agile”, the usual suspect analysts will be predicting the Next Big Thing in business agility, and I still won’t have finished my Software Faster book. And the robots still won’t have taken over. Yet.
ALI: What are you most looking forward to about coming to Agile-Lean Ireland 2019 in Dublin?
Dan: The craic! Seriously, I love spending time in Ireland. I was working with an organisation in Cork for a while who were just wonderful. The Irish have a healthy balance of seriousness and irreverence. Stuff gets done and people manage to have fun doing it. I like being around that.
ALI: What can our attendees expect to hear from your talk at ALI2019?
Dan: This is a collection of things people have said or done at key points through my career that really impacted me. They didn’t realise at the time, and often I didn’t either. Pull up a chair. Crack open a beer or pour a cup of tea. Let me tell you some stories.
Dan will be one of our keynote speakers at Agile-Lean Ireland 2019 at Croke Park, Dublin, on April 25-26th 2019. Don’t miss you chance to see him by getting your tickets now.