In this podcast recorded at Agile 2017, Shane Hastie, Lead Editor for Culture & Methods, spoke Troy Magennis about his talks at the conference on “I love the smell of data in the morning” and “10 ways to choose what to start next”
Key Takeaways
- The absence of data means all you have is an opinion
- The simplest way to start gathering data in software development is to put the starting date/time on one corner, the ending date/time on another corner and use different colour post-its for different types of work
- It’s the trend that’s important, not any one week of data
- Your job is to make better decisions with the least amount of effort
- Context is hard to spot if you are in the middle of it – you need to be able to step away and see from a different perspective
- The job of a leader is to help the people who report to you understand the bigger picture and the impact of their decisions on the whole
Subscribe on:
Show Notes
- 0:20 Introductions
- 0:58 The absence of data means all you have is an opinion
- 1:30 The misconception that you need a lot of data to make any decision at all – seven to eleven samples is sufficient to outdo intuition-based decisions
- 2:03 The simplest way to start gathering data in software development is to put the starting date/time on one corner, the ending date/time on another corner and use different colour post-its for different types of work
- 2:10 This simple set of data is sufficient for forecasting and providing a balanced dashboard of system qualities
- 2:54 The three simple pieces of data can be used to generate a wide range of analytics – Troy produced a spreadsheet with 27 charts from those three simple metrics
- 3:32 It’s the trend that’s important, not any one week of data
- 3:44 Metrics need to be team-level not individual focused – if you start embarrassing people then you will get crappy data
- 3:58 You never show one metric at a time
- 4:13 Help teams to use data to make smart tradeoffs
- 4:20 There will always be a tipping point at which throughput impacts quality – make this visible
- 5:05 The real value in the data is for forecasting and selecting what to do next
- 5:48 Exploring ways that data can be used to help make a sequencing decision by asking what can go wrong with different approaches
- 6:34 There is no one right answer – sequencing decisions are totally context specific
- 7:03 Stop being so dogmatic about ways of making decisions and look at context
- 7:12 Your job is to make better decisions with the least amount of effort
- 7:48 Context is hard to spot if you are in the middle of it – you need to be able to step away and see from a different perspective
- 8:05 Tools for seeing the context – Wardly Maps and Story Maps
- 8:50 Teams are making trade-off decisions constantly – make sure they have the best information to do so
- 9:28 Provide the context information to the teams to help people make the decisions you wish they would have made
- 10:14 Advice to technical influencers/leaders – be wary of the dangers of pushing too hard on any specific metric or behaviour
- 11:05 When faced with good ideas ensure you can identify the impact on other aspects of the work
- 11:10 Often a good decision in isolation is a bad decision universally
- 11:34 The job of a leader is to help the people who report to you understand the bigger picture and the impact of their decisions on the whole
Mentioned:
