Notes from YOW! 2014: Mary Poppendieck on ‘The (Agile) Scaling Dilemma’

Share Button

I attended YOW! Sydney 2014 and thought some people might get something useful out of my notes. These aren’t my complete reinterpretations of every slide, but just things I jotted down that I thought were interesting enough to remember or look into further.

Lots of empty seats at a stadium. Can Agile scale to this kind of crowd?Mary Poppendieck (@mpoppendieck) spoke about scaling agile teams. (Slides)

She started by saying:

“There’s a big assumption that if agile is good, scaling agile must be good.”

Which made my jaw drop. I make that assumption. It had never occurred to me. Maybe agile techniques don’t work in a larger organisation?

She talked about four constraints on scaling: system complexity, organisational mindset, multi-team communication, and the time and energy of bright creative people.

She encouraged the idea of having a Product mindset instead of an IT mindset: we should focus on customer delight and developing profits for reinvestment, rather than delivering projects on time and on budget and reducing costs.

Speaking about organising teams into groups, she suggested that any teams that have a shared system test should be one group.

Following from that, she said if groups are too large, change the architecture so that you can split the groups. She seems to be proposing a reverse of Conway’s Law: change the architecture first, and then change the organisation to match.

“If teams need to work together, they need one goal.”

She said to think about the possibility of using the military model, where leaders have strategic awareness (what the sub-team goals are) two levels down and situational awareness (how things are going with the wider campaign) one level up. If the platoon is failing to meet its goals, the squad (a division of the platoon) has to adapt what it’s doing to help the platoon as a whole.

Stop thinking about delivery teams and have problem solving teams. Work back from the desired impact. Start with “Why?”.

She finished up with these two recommendations:

  • Understand the desired impact of what the group is doing.
  • Measure (continuously) that the impact can be achieved.

Want to learn more?

Image credit: ‘Silent Crowd‘ by Alfonso

Share Button

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.