Multi-learning and # Backlogs by Lv Yi

In the large-scale product development organization, backlogs are pervasive. Backlogs are the manifestation of various specializations. Why more backlogs? Efficiency is the main driver. What are the consequences? The increased end-to-end cycle time for delivery and the reduced flexibility to maximize value. The number of backlogs is the key lever in achieving agility - fewer backlogs, more agility. Multi-learning means learning across functions, technical domains and customer domains. More multi-learning enables fewer backlogs, while fewer backlogs drive more multi-learning.

Join us for this conversational session with Lv Yi as he explores how large-scale product organization grows rigidity and possible ways to deal with it.

Transcript

1. #backlogs & multi-learning
   Lv Yi, Odd-e
   

   View Slide

2. dependent backlogs
   Component A
   Tasks
   Component B
   Tasks
   Component C
   Tasks
   Analysis
   Tasks
   Design
   Tasks
   Testing
   Tasks
   

   View Slide

3. multi-learning for
   functions and components
   • real team with one backlog (i.e. cross-functional feature team)
   
   • specification by example
   
   • collective code ownership
   
   • pair/mob programming
   
   • component mentors
   
   • current-architecture workshop
   
   • multi-team design workshop
   
   • communities of practices
   

   View Slide

4. independent backlogs
   Feature
   Items
   Feature
   Items
   Feature
   Items
   

   View Slide

5. multi-learning for
   customer domains
   • multi-specialized feature teams with one backlog
   
   • multi-team PBR
   
   • initial PBR
   
   • overall PBR
   
   • one sprint review
   
   • review bazaar
   

   View Slide

6. summary
   • the more dependent backlogs, the longer cycle time
   
   • the more independent backlogs, the less adaptiveness
   
   • The ultimate lever is to reduce #backlogs
   
   • multi-learning enables fewer backlogs
   
   • fewer backlogs drives multi-learning
   

   View Slide