Making the Change - Going Agile at the Department of Labor

Making the Change:
Going Agile at the
Department of Labor
Presented by Joey Spooner
Agile 2017
August 7, 2017
Copyright © 2017. TriTech Enterprise Systems, Inc. All Rights Reserved. CONFIDENTIAL INFORMATION.
1

Who am I?
u 15+ years in technology
u Management with a focus towards
small teams
u Product design and development
u Accredited Kanban Trainer
u Kanban Coaching Professional
u Certified ScrumMaster
Copyright © 2017. TriTech Enterprise Systems, Inc. All Rights Reserved.
2

The Department of Labor
3

Background
4
Teams
Division
Office Public Affairs
Electronic
Communications
Technical (4-5)
Governance and
Security (1)
Quality
Assurance (2)

Scope
5
u Open Data API Platform (Quarry)
u Aging Applications
u Helping other agencies
u Performing QA (mobile accessibility)
u Releasing their applications (Google Play, Apple iTunes)

Focus of this talk
u Two periods
u August 2014 through December 2015 – proto-Kanban
u January 2016 through August 2016 – full Kanban
6

Five Questions
u “How can you help me to see into the work?”
u “How long will it take for work to complete once the team starts on a story?”
u “When will a feature or project be done once the team starts?”
u “How can you help me to express to upper management the demands on the
team and their capacity or capability to address the demand?”
u “How can you get QA to talk with the developers?”
7

Stabilizing Scrum (Technical Team Only)
8

9
2
4
4 4
Stories
Application 1
(Low Priority)
Application 2
(High Priority)
Tasks from Stories

Ticket Design and Collective Ownership
10
GitLab Issue ID
Ownership

Working with Batches and Iterations
u Not Scrum, less and more simple
u Prioritized backlog of stories for more than one project
u How many stories do you think you can complete in 3 weeks?
u Number of stories grouped into a batch and iteration began
11

Benefits and Challenges
Benefits
u Clear sense of what was
important, what was committed
for delivery
u The opportunity to reflect as a
whole team via retrospectives
u Making decisions to improve using
data (e.g., cycle time, throughput,
and WIP)
Challenges
u GIGO can happen in any “Agile”
environment
u Broken builds on a small team
really suck
u Batching is a …
12
August 2014 through December 2015

Performance
u Average Cycle Time – 55 days (from selected to ready for delivery)
u Average Daily WIP – 14 items
u Average Throughput - .26 items per day (about 25% of a story per day)
u Average Selection - .25 items per day (25% of a story per day)
13
August 2014 through December 2015

Five Questions: How did we do?
14

Using the Kanban Method
u Why?
u To look into the future (probabilistically)
u How much demand?
u How much capability if we removed the batching?
u Acceptance of last minute requests
15

Developing a Flow Mindset
16

17
Implementation

Tooling Jira
u Changes we made along the way
u priorities to reflect the service levels
u Sub-tasks to operate as “issues” (blocks, defects) preventing work from proceeding
u Transition screens to remind people of their policies for a given column or status
18

Using the Method’s cadences
u Replenishment (scheduled) in place of Sprint Planning, etc.
u Daily planning in place of Daily Stand-Up
u Service Delivery Review in place of Retrospective
u Delivery planning in place of a dedicated Sprint to release software
19

Sizing Challenges
20
By Scarlet23 - Own work. Based on the raster graphic Image:Megalodon scale1.png, as well as Image:Rhtyp u0 white bg.gif,
Image:Carcharodon carcharias01.jpg, and Image:Human.svg., CC BY-SA 3.0,
https://commons.wikimedia.org/w/index.php?curid=25031119

Benefits and Challenges
Benefits
u Limiting WIP across the whole
system (not just for in progress per
area)
u We were able to forecast when
something might be done
u Using the Kick-start Field Guide
made it a lot easier to get going
Challenges
u Ideal design and policies
u Jira has its limitations
21
January 2016 through August 2016

Performance
u Average Cycle Time – 11 days (down from 55 days)
u Average Daily WIP – 4 items (down from 14 items)
u Average Throughput - .18 items per day (down from .24 items per day)
u Average Selection - .30 items per day (up from .25 items per day)
u Does anyone see a problem here? What’s going to
happen in about a week or two or maybe longer?
22
January 2016 through August 2016

Five Questions: How did we do?
23

Lessons Learned
u Leadership is the gasoline
u Big things going into a machine will put more stress on it (minimize the size)
u Stick with the Kanban Method’s principles and practices when the going gets
rough
24

Thank you!
u Joey Spooner
u jspooner@tritechenterprise.com
u @spoonstein
u www.tritechenterprise.com
25

Making the Change - Going Agile at the Department of Labor

More Related Content

What's hot (20)

Similar to Making the Change - Going Agile at the Department of Labor (20)

Recently uploaded (20)

Making the Change - Going Agile at the Department of Labor