Top Key Take-aways from the Agile 2014 Conference

Key Take-aways
Hemant Elhence, CEO
Vinayak Joglekar, CTO
Confidential – August 2014
Conference Overview
• July 28 – Aug 1 in Orlando, FL
• Approx 2000 participants from 40 countries, 17 tracks, over 200
sessions, plus inspiring keynotes
• Stated theme: Achieving Enterprise Agility
• We attended 20 sessions each, plus Exhibit Booths of about 40
vendors of tools and training services
• 4th year of 1-day Executive Forum, with invited senior executives, ~25
• More large company participants, e.g. Walmart, Cisco, etc.
Confidential – August 2014
17 Tracks
• Agile Boot Camp
• Coaching and Mentoring
• Collaboration, Culture &
• Development Practices &
• DevOps
• Enterprise Agile
• Experience Reports
• Leadership
• Learning
Confidential – August 2014
• Lightening Talks
• Open Jam
• Project, Program and
Portfolio Management
• Research
• Stalwarts
• Testing & Quality
• User Experience
• Working with Customers
Synerzip’s Top “10” Takeaways
Note: 40 sessions were attended out of 200 by
Hemant Elhence and Vinayak Joglekar. Where possible,
the presenter’s name and other reference are listed.
Confidential – August 2014
Top 10 (15) Topics
1. VJ: Managing People
2. HE: Scaling Agile/SAFe
3. VJ: Lean Start-up
4. HE: Spotify
5. VJ: Mob Programming
6. HE: Self Organization –
TradeMe Case
7. VJ: T-Shaped + Broader
8. HE: Agile Transformation
Confidential – August 2014
9. VJ: UX Runway +
Mobile UX
10. HE: Value Team vs.
PO Role
11. VJ: DevOps & Continuous
12. HE: Estimation
13. VJ: Org Structure &
14. HE: Interesting Soundbites
15. VJ: Interesting Soundbites
1. Agile is not about software
It’s about managing people
Confidential – August 2014
Managing People
• Healthy, happy workplaces are
real and achievable  healthy
• Trust is the bedrock of high
performing team
Confidential – August 2014
Managing People
Confidential – August 2014
Managing People
• For building trust/commitment, embrace
conflict − think about things never discussed
• “We need to allow people to choose to
unleash their potential” - Olaf Lewitz
• We need to build fail safe relationships
• Very simple manifesto − We value people
• “Self organizing organization” - Sandi Mamoli
Confidential – August 2014
Managing People (Cont’d)
• Theory Y to deal with Business VUCA
• Need based allocation instead of budget to
avoid gaming & conflicts of target/allocation
• No perfect KPI-Standing on scale example
• Roundabouts instead of traffic signal
(Value based system instead of rules based one)
Confidential – August 2014
Traffic Signal Vs Roundabout
Confidential – August 2014
2. Scaling Agile
• Effectively scaling Agile adoption to hundreds of teams
• Probably the biggest discussion topic, perhaps the next
biggest frontier
• Multiple frameworks/approaches
Scrum-of-Scrum (SoS)
Large Scale Scrum (LeSS, Larman/Vodde)
Scaled Agile Framework (SAFe, Leffingwell)
Disciplined Agile Delivery (DAD, Ambler/Lines)
Spotify “Model” (Kniberg)
Scrum at Scale (Sutherland,, meta framework
• Agile Scaling Knowledgebase (ASK) Decision Matrix,
• Less about the framework choice, more about culture change
Confidential – August 2014
Confidential – August 2014
3. Lean startup adoption
Beyond discussion −
GE does it.
Confidential – August 2014
Lean Startup in Microsoft & GE!
• Keynote by Sam Guckenheimer confirmed
Microsoft’s journey to cloud cadence 
customer feedback in sprints from 2008
• Is it important? Going in the right direction?
• Early customer feedback is the primary
motivation for Agile  Cloud Cadence move
• C/s functional teams - autonomous backlog
• Deployment in 3rd week of 4 week sprints
• Only trunk checkin - baby steps - feature flags
• GE Healthcare executive lunch conversation
Confidential – August 2014
Lean Startup meets Agile via UX
• Workshop to quickly design an MVP in a cross
functional team – diverging  converging
• Design:
Thinking + Lean Startup + Agile = Lean UX
• User-centric continuous design vs big upfront
• Hypothesis driven development
• Prototype  Test  Learn cycles
Confidential – August 2014
4. Spotify “Engineering Culture”
Spotify offers a fascinating model for scaling. It has kept an agile mindset despite having
scaled to over 30 teams across 3 cities.
Alistair Cockburn (one of the founding fathers of agile) visited Spotify and said “Nice - I've
been looking for someone to implement this matrix format since 1992 :) so it is really welcome
to see.”
Confidential – August 2014
Squad, Tribe, Chapter, Guild
Squad: the basic unit of development, similar to a co-located Scrum team,
designed to feel like a mini-startup. Each squad has a long-term mission such
as building and improving the Android client, creating the Spotify radio
experience, scaling the backend systems, or providing payment solutions.
Tribe: A tribe is a collection of squads that work in related areas – such as the
music player, or backend infrastructure. Designed to be 100 people or less, a
tribe can be seen as the “incubator” for the squad mini-startups.
Chapter: A chapter is your small family of people having similar skills and
working within the same general competency area, within the same tribe. For
example the testing chapter, the web developer chapter or the backend
Guild: A Guild is a more organic and wide-reaching “community of interest”, a
group of people that want to share knowledge, tools, code, and practices. A
guild usually cuts across the whole organization. For example, the web
technology guild, the tester guild, the agile coach guild, etc.
See this white paper by Henrik Kniberg & Anders Ivarsson
Confidential – August 2014
5. Mob Programming
• Mob programming-Woody Zuil-Teams attracted
by awesomeness of vision=naturally motivated
• Best requirements, architecture and design
emerge from self organizing teams
• 1 computer, 2 keyboards, 2 projectors - bright
minds working on the same thing at the same
time in the same room for 3 years!
• Interactionskindness, consideration & respect
• No politics, context switching or waiting
• Less code, no duplication, low technical debt
Confidential – August 2014
Mob Programming video
Confidential – August 2014
6. Self-organization
• Trade Me case study on self-organization leveraging Spotify WP
• Company ( size 125 people, 80 in engineering
• Facilitated self-selection process, done offsite
Pre-defined 11 Squad categories (product areas)
Pre-assigned Product Owners
Rest of the team of dev & testers self-selected
“Do what is best for TradeMe” – posted prominently
Squads should be 3-7 people
Capable of delivering end-to-end
Co-located, if possible
• Result
– Stable, focused teams, that deliver higher
– Happier team members
• Mob Programming – evolved version from pair programming
Confidential – August 2014
Self Organizing, Self Selecting
• Spotify concept of squads, tribes, etc. was the
inspiration to squadify all teams
• Idea from Fedex day - teams were formed by
self selection - what if they do it all the time?
• 3 out of 11 squads self-selected on trial day
• No explicit management buy-in was asked for
• Empty squad sheets - only 6/7 pictures could fit
• Sheets placed, title/mission explained by PO 10 min iterations  8 squads fully staffed
• Learning - Trust works, peoples’ interests
Confidential – August 2014
7. T-Shaped + Broader Skills
Everyone does everything
Confidential – August 2014
T shaped skills
Every member has QA, BA, Ops & Dev skills
No one works alone; everyone works together
Shared activities  people more accountable
Communication, curiosity, respect, empathy
needed for cross-skilling within the team
• Remove phases & design together; no story
would be written by less than 3 people
• Shared activities resulted in better quality, 0
bugs, 3X efficiency, shared understanding of
goals, trust, increased empathy
Confidential – August 2014
• “Do what you love to do, work with purpose,
care for your tribe” - Diana Larson
• Tools for self-discovery − survey, ask
friends, check what you did when there was
nothing to be done, the kind of stories that
move you
• Purpose inspires, brings focus, motivates
Confidential – August 2014
Lean to smash biases
• Lean canvas - cognitive bias - I know it
• Experiment A3 - hindsight bias - I knew it
• Heaven/Hell ritual - Optimism bias - Wow its
• As the problem gets fatigued disciplined
thinking fast and slow helps
• Your sense of doubting brings disciplined slow
Confidential – August 2014
8. Agile Transformation
• Big challenge, faced by large (>10,000 person), older companies, with
distributed (incl. offshore) teams
• Entire focus of the Executive Forum, with case-studies by: Bank of
America, United Healthcare, Frost Bank (146 yr old), Cerner Corporation
• Culture trumps process and practices
– Agile Mindset
– Servant Leadership
– “Less about the framework (e.g. SAFe), more about the culture change.”
• Customized Agile process, with some remnants of waterfall
• Embedded coaches
– Hired externally, but integrated
– 1 coach per 3-6 teams
• Physical space changes – cubes to open room, all whiteboard
• Lot of attention to communication of vision, why change, etc.
• Attention to metrics
Confidential – August 2014
9. UX Runway + Mobile UX
Integrating UX design
& scrum
Confidential – August 2014
Intro to the UX world
• Roles − researchers, information architects, visual
designers, CSS devs, accessibility experts
• Aversion to time boxing; Centralized / Isolated
• UX needs to be integrated at portfolio, program and
product levels; hence it’s hard to integrate
• User research/personas etc. at portfolio level
• CSS designs/wireframes at product/program level
• UX runway with + 3/2/1 week lead for research,
architect and dev work
Confidential – August 2014
Step by step UX framework
• Takes 2-4 weeks before development
• Collaborate/empathize with users for
problem/assumption validation
• Business goals – why are we doing this?
Expectation setting from business owner
• Contextual research – observe users in their
• Target user may change as we iterate; but
its important to know the primary persona
• Research insights = opportunity patterns
Confidential – August 2014
Step by step UX framework
• Lean canvas or elevator pitch to prioritize
• Risk/Complexity/Value to prioritize
• Design principles are best understood by designers –
rest solution can be framed by team
• Experience principles like “proactive, flexible, friendly,
safe” need to be explicitly stated
• Storyboards/User Journey – Day in the life
• Lowest fidelity proto to get early evaluation
• Experience mapping across various touch points
• Interaction/navigation model – core behavior
Confidential – August 2014
Mobile UX testing
• UX research needs to be planned with dev
• Heuristic review – is the design conforming to
standards and known rules?
• Usability study – in-person/remote
• Remote research is difficult for mobile apps
• Working with agile teams need UX to be flexible on
fidelity of prototypes – could be paper or using tools
like appcooker/blueprint
• Low budget research – observe users – don’t talk
• Magitest/remoteviewer – record/share remotely
Confidential – August 2014
Mobile UX considerations
• Case study – how to display menu; 9 out of
12 didn’t know the hamburger icon
• Secondary navigation – accordion preferred
• Horizontal scrolling was understood when
the arrows appeared in the table header
• 4X more appetite for scrolling on tab vs
Confidential – August 2014
10. Value Team vs. PO Role
• Value Team session by Ahmed Sidky,
• Considering Product Owner as the single wring-able neck for feature
prioritization is unreasonable
• Create a Value Team, PO = Value Team Facilitator
Confidential – August 2014
11. DevOps & Cont Delivery
DevOps is more about
culture than tools
Confidential – August 2014
DevOps-Next level of c/s
DevOps isn’t @ creating a silo or buying a tool
DevOps is a cultural & professional movement
Can’t solve social/cultural issues with tools
More and more sysadmins are writing code
Many devops - day talks - empathy for other depts
Continuous change often gets misunderstood as
nothing’s ready because it's going to change
• Small teams that are inquisitive & ready to learn
• Hackathons − build anything that added value
• ^ pressure, ^features, v design  Sinking of Vasa
Confidential – August 2014
12. Estimation Evolved
#NoEstimate Movement
• “Building software is by its very nature unpredictable and unrepetitive. While
building software we cannot easily break down the work into same-sized,
repeatable widgets like we can when manufacturing car parts.”
• Just deliver high-value working software slices in shortest possible time
intervals, and do-away with estimation.
Using Empiricism over Guesswork
• Use actual historical data of work done by team
• For large projects, sampling based estimation
– Take a sample of few epics to use
– Use Weibull distribution, with Monte Carlo simulation tool for arriving at full project
• [email protected]
Confidential – August 2014
Using Sampling + MC
From Troy Magennis’ session Moneyball for Software Projects
Confidential – August 2014
13. Org structure = Architecture
Designs = copies of
communication paths
Confidential – August 2014
Architecture & Org Structure
• Architecture originates in different places and
different ways, and is deeply ingrained in the org
• Shape of design group = imp design decision
• Loosely coupled/distributed orgs are less likely to
produce tightly coupled ball of mud
• Selective hiding and exposing information in an
org − like interfaces with private/public methods
• Architectural change required for remote team
• Avoid hierarchy paths, pipelines; refactor org
• Scaling-^ small teams building PAAS services
Confidential – August 2014
Interesting Soundbites (HE)
• “Agile” is technology driven, Lean-Agile is business value driven
• Agile helps you deliver faster, but doesn’t save you from bad business
• To pilot Agile, pick a really important (& visible) project in the company
• Keep all the work visible, including any skunk works, pet projects, etc.
• To really understand customer needs, don’t (just) ask them; observe
them, and infer what they really need
• 2 key requirements for Agile to not fail:
1. Definition of Done
2. Culture
• Candidates for 5th Value for Agile Manifesto
– Demonstrating & delivering business value OVER trying to use the word Agile to sell
– Continuous improvement OVER “Best Practices”
Confidential – August 2014
Interesting Soundbites (VJ)
• If we improved by measuring we could’ve reduced weight
by just standing on the scale.
• Co-located team is likely to result in a local mudball.
Remote teams result in a modular approach.
• Architecture is a social discipline. It impacts daily life.
• Many scrum teams work in a compressed waterfall model.
• By asking users to participate in early research you are
getting their trust for the brand not damaging it.
• Politics tax is the amount of time spent to CYA.
• The very word “engineering” is limiting our thinking of what
software can do.
• Continuous delivery is a journey. No need to get there now.
The direction is right.
Confidential – August 2014
Confidential – August 2014
Confidential – August 2014

similar documents