Agile Australia 2013 was recently held in Sydney with over 850 attendees and 3 days. Between running a pre-conference workshop, recording interviews for InfoQ, presenting a session and being a MC for a number of sessions, it was a fairly busy time but I did get to sit in on a couple of sessions.
I was once again one of the conference advisors, although this year we introduced the role of Stream Chairs and Reviewers who took the bulk of the review of the 240 plus submissions and, I think, went some way to making the whole selection process more transparent and community driven.
Pre-Conference Workshop – Introduction to Agile
There were a number of pre-conference workshops running on the day before the conference, and on behalf of Software Education I ran an Introduction to Agile workshop for a small but engaged group of people new to Agile. According to the course overview:
This course provides an independent one-day introduction and overview of Agile Software Development. We look at the underlying philosophy and motivation for this trend in software development and examine the core values, principles, practices and techniques that fall under the broad “Agile” umbrella. Independent of any single brand, this course looks at the key factors that are needed to apply Agile effectively and provides an experiential introduction to working this way.
InfoQ was a media sponsor for Agile Australia this year, and being the Australian based Agile Editor for InfoQ, I undertook the organisation iof the recording of sessions and interviews. I recorded a number of interviews throughout day 1 of the conference and I look forward to seeing them available on InfoQ in the coming months.
Visual Management: Leading With What You Can See
The session I presented with Renee Troughton had a great turnout and plenty of questions afterwards. The slides are available in a separate post.
In relation to the sessions that I attended, here are my notes.
Keynote: The Lean Mindset
Mary Poppendieck delivered the opening keynote, her slides are available here.
- going beyond software
- Thinking Fast and Slow by Daniel Kahneman – two types of thinking – system 1 (fast, instinctive and emotional) and system 2 (slower, more deliberative and logical)
- need to see the waves and learn how to surf them, rather than trying to control the waves
- cognitive biases – confirmation biases (align with people with same thinking), anchoring (hang on first piece of information), loss aversion (careful not to lose what we already have rather than gain)
- dealing with options – teenage decision making (either/or or whether or not), widen the frame (what about both or none of the above), multiple options, patterns
- speed and quality are compatible, you learn that you can’t go fast without high quality
- Continuous Delivery by Jez Humble and Dave Farley has become the bible, supersedes iterations and Kanban because it’s too fast, whole feedback loop from customers to the development team
- Ericsson – the past is not good enough for the future, stop projects because they are big batches
- accept uncertainty and learn how to live with it – range estimates over point estimates and manage flow
- managing complexity – probe, observe and adjust; dealing with it – flow, obstacle, adjust
- wisdom of the crowd, widen the perspective, zoom in/it, look at base rates (probability of success in your market and what makes you different)
- why do companies fail – they focus too much on success, take what they have and make it better, forget to look at the big picture and see what the world is doing – Garmin and Magellan lost 70% of their market overnight due to Google Maps on the iPhone
- Carol Dweck on the Growth Mindset – fixed mindset versus growth mindset
- perfection paradox – learn to fail and learn to prevent rather than striving for perfection, resilience is learning to fail
- flow depends on your engagement – deeply engaged, no distractions, time evaporates – otherwise you drift between anxiety and boredom depending on your expertise
Keynote: Managing for Serendipity
I was quite looking forward to seeing Dave Snowden and hoping his talk would cement in my mind the Cynefin Framework. His slides are available here.
- it is better to have a partial view of the whole, than a complete view of the parts
- theory deals better with uncertainity than practice
- exaptation (taking something that exists and enable it to a different purpose, usually two unrelated things)
- agile manifesto swung too much towards customers and away from how we can educate them
- you absorb complexity, don’t delude yourself into thinking you can eliminate it
- outliers are where opportunity and threat manifest themselves first - most research and search eliminate them
- only way you can understand a complex system is via experimentation.
- evidence supports competing hypotheses
- Dave’s Harvard Business Review article
- you want to cycle between complicated and complex, stuff that goes to simple usually becomes complacent and dies
- cynics care about your organisation – they are the only ones making noise
- meetings should not be used to resolve problems – you get dominant players – you want to to do 5-7 parallel safe to fail experiments
- interventions should be coherent, safe to fail, finely grained, tangible – tackle problems oblique (plant different ideas, like managing a teenager), naive (anthropology experts in the Holiday inn), a few high risk high return options
- process is what you need to change people (sorry mainfesto), naive if you think you an change one person at a time
- complex domain action form – if it doesn’t fit in top 4 boxes it is not safe to fail
- micro narratives are what we want (stories developed in workshops are not the same as the discussions around the water cooler)
Keynote: Beyond Budgeting
The highlight of the conference for me was to meet and hear from Bjarte Bogsnes, as I have long been a fan of his Beyond Budgeting work. His slides are available here.
- the purpose is not to get rid of budgets, need to change traditional management and the budgeting mindset to make organisations more agile
- written into The Statoil Book, but has not reached all corners of the organisation
- transparency is a great social control mechanism and good for learning
- management recipes – someone has done all the thinking for you, beyond budgeting is not that
- Beyond Budgeting principles
- we budget for targets, focussing and resource allocation – inefficient and in one number – separate the numbers and be event driven not calendar driven
- change the mindset from do I have budget to is this really necessary and how much value it is creating
- Statoil abolished annual budgets, but they still have project budgets but can get money at any time (the bank is always open)
- need good information to make decisions – current status as well as capacity
- use a burn rate rather than a budget, with some variance
- use a unit cost, keep exploring as well as you stay within a unit cost
- bottom line – good costs create value, eliminate the bad ones
- express cost in words rather than dollars as strategic actions and objectives
- need to start trusting our people
- “the way we deliver is as important as what we deliver”
- objective > measure > actions > goals
- KPIs – perfects KPIs do not exist, make them ambitions to action
- dynamic forecasting – don’t put everybody into the same limited time buckets – some people need small time frames, some people need years
- apply pressure to the KPIs – were they ambitious, did you stretch versus low balling, were there outside influences
- mechanical link and no judgement between KPIs and performance bonus is dangerous
The Guessing Game: Alternatives to Agile Estimation
Neil Killick has become the voice of the #noestimates discussion in Australia (whether he likes it or not). His slides are available here.
- estimation – what am I going to get and when
- estimates set an expectation level
- in 12 moths time if estimates come true – either by good guessing or by adjusting for over runs
- we have known unknowns – can’t predict them and can’t ignore them
- use real constraints – what can we build for this money, budgets create a real deadline and bring out creativity
- create mini constraints and iterate and learn, come up with a mini solution for what we can do, small iterations to keep options open and diversify our risk
- need an a-team that can build solutions continuously
- iterative pricing – allows the customer to cut the cord early, possible even with traditional contracts
- present customers with expectations of of what they are going to get and when – flexible options
- slice things to the same size, avoid story points, count cards – price per feature
- story points will be gamed – people will make the burn chart look good
In Conversation with Patrick Eltridge
I was the MC for this session that was a coffee table conversation between Beverley Head and Patrick Eltridge, the CIO of Telstra. When I introduced this session, I made the comment that it was interesting to see how Telstra was progressing on their Agile journey and Patrick was at the conference for his third year now; in the first year we didn’t really believe they would be able to make an Agile transformation and in the second year we weren’t sure how much was fact and fiction. In 2013, they are certainly making their presence felt with over 70 people at the conference, a title sponsorship and a number of sessions being presented.
I did not get to hear all of the session, but hear are some snippets I picked up
- IT strategy needs to be driven by the corporate strategy, then creating an environment that can change and people being the best they can be
- huge “opportunity” when he joined Telstra
- celebrate successes as much as possible – stiffens the spine for the naysayers
- reverse mentoring – mentor older staff with younger ones to pass on new thinking and ideas, both sides learn
- nations can benefit if agile leadership is successful
- simplified scorecards and KPIs – single number of financial performance (EBIT), engagements with stakeholders, employee engagement plus project outcomes
- Scaled Agile Framework and weighted first has allowed Telstra to have acute business control and adapt as required
This discussion also spawned some news articles around graduates mentoring CEO’s and five day interviews.