Inspect and Adapt Then What? @coreyhaines
Agile!
Agility!
Agility Feedback
Agility Minimize time between Doing Something
Agility Finding out if it was the Right Thing To Do
Agile Methodology (XP, Scrum, etc) vs Agility
Inspect and Adapt
Inspect and Adapt Fixed Methodology and Context-Specific Practices
This talk Example from my last two companies Different Techniques Different Contexts
ME! That’s Squeaks ——>
Me! • Software Developer • Introduced to Extreme Programming (XP) in 2004 • Been running that way ever since • Currently CTO of Hearken • Does not have the answers, only experiences • Loves Cats!
Examples
Planning • Two companies: PowerReviews and Hearken • Different Contexts
PowerReviews • Small, 20-person company • 5 different stakeholder sections • 3 developers • Single platform
Planning Problem #1: Estimation • 3 developers • Only one familiar with codebase • One developer very junior • How do you estimate?
Wrong Question How do you estimate?
Why must we estimate?
What value do we get from Estimation?
Deadline vs Scope
Replace with small, similar tasks by splitting feature?
Focus on scope sequence appropriately
Often don’t have to do everything
Prefer Removing Something If find struggle or problem with process
Planning Problem #2: Prioritization • 5 stakeholder sections for 3 products • Everyone votes for their own cards • Loudest person gets their cards selected
Sales Content Prioritization Queue Moderation Technology Data
Priority???
Value???
High Value Value to Company Low Value Courtesy Arlo Belshi Time Quick To Do Long To Do
Value-Based Sequencing • Based on Financial Value • Money Earned • Money Saved • Sequence by value
Example: Fraud Flags • Moderators look for fraudulent reviews • 4 steps • Click name to get IP • Navigate to IP Search Page • Enter IP • See how many / frequency of reviews from IP
Hard to get this prioritized over Sales team requests
Value Calculation • Have to do for every review • 20 seconds per review • ~ 200 per day • (20 sec / review * 200 review / day ) ~ 1 hour / day • 75 reviewers = 75 hours per day • $20 / hour * 75 hours = $1500 / day
Got Sequenced!
This can’t work! • Time-boxed experiments • Let’s try for 2 months • Put meeting on calendar to address whether this works
Hearken Tiny, 9-person company (started at 3, now at 9) 1 stakeholder section for single product 3 developers
Planning • Start with no estimation • Start with weekly planning meeting • Stakeholders sequence tasks
Problems • Value??? • Money? Hard to gauge. Early-stage. • Save time for customers? Hard to gauge. Early- stage. • Sequencing becomes rubber-stamp • Rarely changes order of sequence • Take time from everyone’s day
Communication
Communication • Small team => Fairly regular communication • “Story cards as placeholder for conversation” • Each story card is defined at time of draw • Expect that story card is split into more than one, put new cards into backlog
Weekly Planning • Present suggested sequence • Anyone have objections? • Done!
New Product Weekly Planning • Building new product • Lots of features to build • Sequencing difficult as no context • Regular “showcase” demos of progress • Using new language/framework (Elm)
Script-based Planning Create script for the showcase What does the product currently do? What doesn’t it do? Work until it can do what it doesn’t do.
Return to story cards after release of product
Context Matters You are special Time-based experiments Frequently change
Inspect and Inspect and Adapt Adapt Inspect and Adapt, Then What? Then What? @coreyhaines @coreyhaines Buy my book! https://leanpub.com/4rulesofsimpledesign/
Recommend
More recommend