Agile Australia 2010 was held for its second year earlier this month at the Crown Conference Centre in Melbourne.
Once again I was honoured to be offered an opportunity to present, but I also had the pleasure of being a conference organiser, an MC for speaker sessions on both days, the opportunity to be part of panel discussion as well as running a workshop with Phil Abernathy on the day prior to the conference for those new to agile. The conference attracted 450 attendees from 140 different companies.
One of the difficulties of assisting in the organisation of a conference like this (and I had a similar experience co-producing a stage at Agile 2010) is narrowing down the large number of proposals into a program that will meet the expectations of the intended audience. There was much discussion and debate throughout the process, but I think we ended up with a well-rounded program that covered a wide range of topics, experiences and session types. I owe a huge thankyou to Rachel Slattery from Slattery IT for giving me the opportunity to have a greater involvement this year and the ability to work with the other organisers Phil Abernathy, Adam Boas, Keith Dodds, Martin Kearns, Dave Thomas and Zaz Teoh.
My notes and summaries are below:
Pre-Conference Workshop: First Steps In Agile
I ran a pre-conference workshop for agile beginners with Phil Abernathy entitled First Steps in Agile, which was based on some of the introductory material that is available in the courses offered by the Agile Academy.
In 3 and a half hours we ran through the fundamentals of agile, and had them creating focusing questions, success sliders and story cards before getting them to play with Lego to deliver an agile project.
Executive Breakfast: Accelerating Agility for Executives and Managers
Jim Highsmith delivered this breakfast seminar to special guests and conference speakers.
- there is no more normal
- efficiency vs responsiveness – which one is your model – Walmart vs Google
- business agility – a key resource is David Sull – The Upside of Turbulence
- if agility is a key bus driver, create a Chief Agility Officer
- there are a number of executive levers
- quality matters, Michael Mah from QSM is a good agile data source
- adaptive quality – can you deliver good quality code in the future, key job for executives is to champion a technical debt reduction strategy
- do less – reduce work in progress, requires a lot of discipline
- eliminate marginal value – Standish have found that 64% of features are rarely or never used
- motivate – autonomy (self organizing teams), mastery (getting better at what we do), purpose (working on value)
- doing agile vs being agile – difference between half assed and half done, executives that not only support agile but understand what it means and be part of it
- “Be the change you wish to be in the world” – Gandhi
- agile leaders – it’s OK to be wrong but not OK to be uncertain, need to balance between up-front and just-in-time
- focus on value not the constraints, what is the capacity to release
- agile 101 – need to start with the rules based prescription, but you can’t get stuck there – move from prescriptive agility to adaptive / adult agility
- salesforce.com – 60,000 tests, all one code base trunk
- all agile models are flawed because the world is a complex place – you need a hybrid approach
- quality is the key to future value
- “It’s kind of fun to do the impossible” – Walt Disney
- Martin Fowler said accumulation of technical debt comes in weeks not months, observed over time that it slows you down
Keynote: Beyond Scope, Schedule and Cost – Optimizing Value
Jim Highsmith delivered this keynote, the slides are available here.
- there is no more normal, it is all change – important to adapt to that change rather than following a plan
- 95% removal of defects on a project is where benefits accrue – good companies in the US average 75% so we have a way to go
- focus on speed and you will get low quality and low speed, focus on quality and you will get speed
- a doubling of staff traditionally quadruples defects, so need to focus on quality
- error feedback – errors put into system when fixing other errors
- create opportunities by doing less – scope is a terrible measure of progress
- do most valuable first and choose the right cut off point, most value for least cost
- value engineering – estimate in both cost and value – calculate cost from the bottom up and value from the top down
- say to product owners – “if you don’t have time to estimate value, we don’t have time to estimate cost”
- process maturity models only look at the process and not the value
- parking lot diagram (FDD) to show what is valuable – only goes green when it is deployable – done done
- value and priority are different – sometimes you may need to mitigate risk or deal with dependencies
- may want to prioritise technical debt stories different to customer stories
- every iteration is the product releasable, focus on releasability and does it meet the vision
- agile triangle – value (releasable), quality (reliable, adaptable), constraints (cost schedule scope)
- being responsive – product, adaptive people, agile process, architect with a vision
- strategic questions – can we release today, value-cost ratio, product quality, within acceptable constraints
- getting funding – business cases still very traditional, need to look at funding models and change them over time, find ways to capitalise expenses or deliver early to get the attention of the CFO – ask CFO what appeals to them about agile
Keynote: Delivering Business Value with Agile
Jeff Smith from Suncorp delivered this keynote. I was pumped because he mentioned my name at least twice!
- agile – fundamental change in how work gets done
- good is not good enough, great is not good enough, only being the best is good enough
- change has always been there, but it is quite disruptive right now
- need to change people from thinking that we can’t do things – what can we do?
- don’t need to change your strategy all the time – you need to adapt it to get value but keep your values consistent
- attract, develop and retain good people is the role of good leaders – job is not to motivate, nurture their passions
- shift your thinking – we don’t know best, look at your partners and competitors
- course correction over perfection – it’s OK to make mistakes
- simplify the way people work together to simplify output
- trust and respect the key values – trust takes character (integrity and intent) and competence (capability and results)
- fear of failure is actually masking the fear of criticism
- a great culture – people love what they do, love who they work for (people leave leaders not companies), love who you work with
- collaboration – contribution, good instruction and feedback
- best thing about Yammer is that it is a flat communication structure – increases productivity and removes cascading messages
- strive for level-less leadership – wisdom of the crowd, removes the communication waterfall
- need coaching, like sports teams, to get better
- constraint is a leaders ability to communicate and direct change, not the business ability to absorb it
- Suncorp is twice as productive as they were a year ago – measure performance by the number of productive business changes into production
- inspire people to flourish personally as well as professionally
- need to break down hierarchies, constraint of thought is caused by management
- remove fear about changes by involving people in the process
Old Tricks for New Dogs and New Tricks for Old Dogs
I was asked to to be MC for this session, and had the privilege of introducing Shane Hastie from Software Education. His slides are available here and you can also download his Project Toolkit Matrix.
- dudes law – value = why / how (David Hussman)
- context – sociology (trust, safe to fail, communication), organization, geography
- babies and bathwater – focus on which one it is, what is of value and what can go down the drain
- all the old tricks still have value
- new tricks – value stream mapping, story mapping, user stories (fundamental agile but may not be the right tool), BDD, pragmatic usability (everybody’s responsibility), paper prototypes (don’t craft a long term artifact, craft for conversations)
- agile constants – TDD (every project should do it), refactoring, continuous integration, automated testing, pair programming, sustainable pace, retrospectives – these are just the givens
- introduced a matrix on how to select the right tools (see link above), and encouraged feedback
It’s the Culture Stupid
I also had the privilege of introducing Rob Thomsett, who’s talk was one of the highlights of the conference for me, his slides are available here.
- modern management which dates back to late 19th century, has reached the limits of improvement
- companies make cultural change because they are hurting or need to – explains hotbed of agile in Europe
- most high level executives believe projects take too long to deliver, project management techniques are bureaucratic
- watermelon projects – “green on the outside, red on the inside”
- don’t under estimate cultural change when you go agile – will make you feel naked
- welcome to agile, welcome to a battle – the war has been raging for 40 years
- outsourcing as payback (referenced in Radical Project Management) – issues is there is no payback
- culture is about values – it’s about what you believe in, it’s sucked into you through life
- standing over somebody and asking for an estimate is an ambush – then play guess the number, price is right, doubling and halving game
- Estimation Games – most downloaded article on the website, but is has made no difference
- Fubini’s Law – every agile person should read it
Kane Mar presented this session, his slides are below:
- all started with continuous integration – the best paper is the original whitepaper by Martin Fowler
- continuous deployment is continuous integration on steroids
- who’s doing it – kaChing, Pirum, imvu, WordPress, Flickr, wiredreach
- continuous deployment gives you quick feedback from customers, avoid waste (reduce cycle time)
- easier in some environments such as web, but it is possible everywhere
- requires investment in infrastructure and testing
- adoption – get continuous integration, stop the (commit) line, simple deployment, realtime alerts, test!
- avoid allowing broken tests, unrealistic expectations
There were a number of scheduled lightning talks, followed by some ad-hoc talks which I unfortunately missed (including one from my colleague Rene Maslen). From the couple of talks I saw:
Big A vs Small a (Tania Broome)
- little a is about delivery, philosophy, set of principles
- big a is about empowering teams and giving them the tools to get there, it’s a statement of commitment to change
- shi, ha, ri – practice the rules, question them then transcend the rules because it is who you are
- big a agile is shu, risk that we stop here
Agile at Home (Ben Arnott)
Ben is another one of my colleagues, and he presented a great talk about how he uses agile at home with his family.
Building an A-Team: I Love It When A Team Comes Together
The session I presented got an awesome turnout (standing room only and people sitting at the front around the stage)! The slides are available in a separate post as well as here.
I wish somebody had got a photo!
I spent a great deal of time during sessions and breaks talking to many different people about a range of many different things. In particular, I spent the majority of the last session having an awesome chat about agile testing with Veenu Bharara from Atlassian.
ThoughtWorks hosted an open house in their Melbourne office after the last session which again was an opportunity to speak with far too many interesting people to mention.