Context Adaptive AgilityContext Adaptive Agility Managing Complexity and UncertaintyManaging Complexity and Uncertainty
Todd Little
2
Are you in the right place?Are you in the right place?G
eek
Hu
mor
2x2 Matrices
Hate
En
joy
Enjoy
Hate
Hopefully the Matrices are funny enough
Enjoy the show!
OK. You don’t have to laugh
Sorry, just content
It Depends
3
The Agile Project Leadership Network Declaration The Agile Project Leadership Network Declaration of Interdependence (www.apln.org)of Interdependence (www.apln.org)
We increase return on investment by making continuous flow of value our focus.
We deliver reliable results by engaging customers in frequent interactions and shared ownership.
We expect uncertainty and manage for it through iterations, anticipation and adaptation.
We unleash creativity and innovation by recognizing that individuals are the ultimate source of value, and creating an environment where they can make a difference.
We boost performance through group accountability for results and shared responsibility for team effectiveness.
We improve effectiveness and reliability through situationally specific strategies, processes and practices.
David Anderson, Sanjiv Augustine, Christopher Avery, Alistair Cockburn, Mike Cohn, Doug DeCarlo, Donna Fitzgerald, Jim Highsmith, Ole Jepsen, Lowell Lindstrom, Todd Little, Kent McDonald, Pollyanna Pixton, Preston Smith and Robert Wysocki
4
Declaration of Independence from Bureaucratic Declaration of Independence from Bureaucratic Project ManagementProject Management
When in the Course of project events it becomes necessary for Project Teams to dissolve the political bureaucracies which have burdened them, a decent respect to the opinions of mankind requires that they should declare the causes which impel them to the separation.
We hold these truths to be self-evident, that all projects are not created equal, that they are endowed by their creation with uncertain and complex characteristics. That project teams are most effective when they value Life, Liberty and the pursuit of Happiness.
5
Penal Management InstitutePenal Management Institute
Now that I am a Penal Management
Professional I can show them how to improve
these Convicts’ Maturity Model
6
7
Hurricane RitaHurricane Rita
8
9
Landmark sells shrink wrap software for Oil and Landmark sells shrink wrap software for Oil and Gas Exploration and ProductionGas Exploration and Production
Users are Geoscientists and Engineers
Subsidiary of Halliburton Energy Services
Integrated suite of ~60 Products
~50 Million lines of code
Some products 20+ years old
80+% of project team stays on same product
Common Model Representation
Well data
Production data
Seismic data
Velocity data
Reservoir /Fluid data
Structural /Stratigraphic data
Common Model Representation
10
We have the challenge of integrating software We have the challenge of integrating software developed in Houston, Austin, Denver and Calgarydeveloped in Houston, Austin, Denver and Calgary
Plus Islamabad
11
Our team was tasked with understanding and Our team was tasked with understanding and improving our software development processimproving our software development process
Maximize our Value Delivery
Scaling to project conditions
Manage Uncertainty and Complexity
Organizational commonality without being overly prescriptiveCore processes
Adaptive processes
What is “barely sufficient?”
12
We were influenced by a number of Agile prophetsWe were influenced by a number of Agile prophets
HighsmithAdaptive Software Development
CockburnCrystal Methods
Boehm and TurnerBalancing Agility and Discipline
Microsoft Solutions Framework (MSF)
13
Manifesto for Agile Software DevelopmentManifesto for Agile Software Development
We are uncovering better ways of developing software by doing it and helping others do it. Through this work we have come to value:
Individuals and interactions over processes and tools Working software over comprehensive documentation Customer collaboration over contract negotiation Responding to change over following a plan
That is, while there is value in the items on the right, we value the items on the left more.
Kent Beck, Mike Beedle, Arie van Bennekum, Alistair Cockburn, Ward Cunningham, Martin Fowler, James Grenning, Jim Highsmith, Andrew Hunt, Ron Jeffries, Jon Kern, Brian Marick, Robert C. Martin, Steve Mellor, Ken Schwaber, Jeff Sutherland, Dave Thomas
14
We studied our project portfolio and identified We studied our project portfolio and identified several project and team attributesseveral project and team attributes
Project ComplexityTeam sizeMission criticality Team locationTeam capacityDomain knowledge gapsDependencies
UncertaintyMarket UncertaintyTechnical UncertaintyProject DurationDependents
15
We scored the Complexity Attributes on a scale We scored the Complexity Attributes on a scale from 1 to 10from 1 to 10
Attribute 1 10
Team Size 1 100
Mission Critical SpeculativeSafety Critical with
significant exposure
Team Location Same Room Multi-site, World Wide
Team CapacityEstablished Team of
expertsNew team of mostly
novices
Domain knowledge gapsDevelopers know the
domain as well as expert users
Developers have no idea about the domain
Dependencies No dependenciesTight Integration with
several projects
16
And we scored the Uncertainty AttributesAnd we scored the Uncertainty Attributes
Attribute 1 10
Market UncertaintyKnown deliverable, possibly
defined contractual obligationNew market that is
unknown and untested
Technical UncertaintyEnhancements to existing
architecture
New technology, new architecture. May be
some "R"
Project Duration 1-4 week 24 months
Dependents/ Scope Flexibility
Well defined contractual obligations or Infrastructure
Independent
17
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0
Project Complexity
Un
ce
rtai
nty
We cross plotted the results and divided the chart We cross plotted the results and divided the chart into four quadrantsinto four quadrants
ixComplexity 10log2 iytyUncertain 10log
2
18
This looked a lot like the Boston MatrixThis looked a lot like the Boston Matrix
Boston Matrix
Market Share
Ma
rke
t G
row
th
Low
High
Dogs
Cash Cows
StarsQuestion Marks
Low High
???
19
So we named it the Houston Matrix and gave animal So we named it the Houston Matrix and gave animal names to the quadrantsnames to the quadrants
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0
Project Complexity
Un
cert
ain
ty
Simple, young projects. Need agilityTight Teams
Skunks
Dogs
Complex, mature marketNeed defined interfaces
Cows
Bulls
Agility to handle uncertaintyProcess definition to cope with complexity
laissez faire
Colts
20
We looked at an overall process flow and how it We looked at an overall process flow and how it would be adaptable.would be adaptable.
Adaptive ActivitiesInputs
Pre-conditions
Project Sanction
RTM
Outputs
Post-conditions
•Released Software
CORE Activities
Iterations
21
We Identified a set of Core Processes for all We Identified a set of Core Processes for all projects to useprojects to use
Aggregate Product Plan
A/B/C List
Quality Agreement
Continuous Integration
Expert User Involvement
Project Dashboard
22
The Aggregate Product Plan sets the high level The Aggregate Product Plan sets the high level vision and expectationsvision and expectations
Project: OpenWells Davenport Project Code: 010265
Product: OpenWellsTarget Date: 3/30/2004
Version: 2003.11.0.0 Release Date: 3/31/2004
Product Manager: Marcus Ridgway SDD: David Field
Vision:Version 2.0 of the new Well Operations reporting and data analysis application. Will bring powerful new query, graphing and reporting capabilities. Comprehensive D&WS input data and output reports will be supported including integration to Production suite.
Platforms:Windows 2000 /Oracle 8.1.7Windows XP / Oracle 9i Windows 2000 & XP /MSDE
Features:18 additional reportsAddtnl apps - Data Anlyzr, NG Profile, AutoprintExtended Rig Equipment supportKnowledge Management - Technical limit drilling,lessons learned, non-productive time, and equipment failuresApplication enhancements (spreadsheet support andtailored well services tab and others)
Strategic Fit:IntegrationWorkflow ( Prototype, plan, actual)Top quartile technology
Target Markets:Existing DIMS customersUS IndependentsNOCsGovernment and regulatory organizationsCompanies requiring integrated offering w/decent wellbore schematic requirementsService companies
23
The A/B/C List sets proper expectationsThe A/B/C List sets proper expectations
A MUST be completed in order to ship the product.
B SHOULD be completed in order to ship the product.
C MAY be completed prior to shipping the product if time allows.
Only “A” features may be committed to customers.
“A” features must fit in a p90 confidence schedule. No more than 50% of the planned effort can be allocated to “A” items
24
A
A/B/C ListA/B/C List
B
C
A
B
CD
50
%
10
0%
Backlog Plan Typical Delivery
50
%2
5%
25
%
25
A/B/C ListA/B/C List
P10 P50 P901 A Honor common login Recognize and support the EDM common login information 5 10 20 M
2 A Certify against old data model
Formally test the new ARIES product against the old ARIES database. If successful, this will ease the transition by giving customers flexibility and enabling them to easily evaluate acquisition databases 10 20 40 L
3 B User group requestsImprove decline calculations to a reserve limit when using hyp/exp switchover 15 30 60 L
4 B Fill RMS functionality gapsProvide multi-level approval and freeze process. Allowing multiple sets of reserve values. 30 60 120 XL
5 C Fill RMS functionality gapsCurrent reconciliation is by EIA or SEC change codes. Need ability for more detail and or user defined arrangements. 35 70 140 XL
6 C User group requestsSTART date with option for a variable DELAY/SHIFT period (number of months) including use with ENDDATE 5 10 20 M
7 C User group requestsExpand range in which we solve for an unknown decline rate in a hyperbolic equation 5 10 20 M
Oxbridge - P&E Systems Requirements
ARIESEstimates
# Rank Name Notes/Purpose/Description T-Shirt
26
We use a Quality Agreement similar to ThomsettWe use a Quality Agreement similar to Thomsett
Attribute
“A”
Very Important
“B”
Important
“C”Not Very Important
Completeness of Functionality X
Completeness of Testing X
Reliability X
Performance X
Installation X
Usability X
Integration X
On Line Help X
Training X
27
Simple, Certain projects are like DogsSimple, Certain projects are like Dogs
They can be trained to be well behaved.
Just the Core practices
Laissez Faire…Similar to Crystal Clear
28
Simple, Uncertain projects are like ColtsSimple, Uncertain projects are like Colts
Young projects with lots of unfettered energy.
Landmark has had success with XP in this quadrant
Short iterations
Daily stand-up meetings
Automated unit tests
29
Complex, Certain projects are like CowsComplex, Certain projects are like Cows
Often are corporate “Cash Cows”More rigorous requirements management; i.e.
use a requirements tool.Requires the generation of useful documentation
for interface definitionsProjects broken up into subprojects and
coordinated by a team of leaders or a Scrum of Scrums.
Project plans with dependency and critical path identification
30
Complex, Uncertain projects are like Complex, Uncertain projects are like BullsBulls
Large projects, hard to control.Need agility to steer through the uncertainty, and
some rigor to manage the complexity.A careful union of the Colt and the CowRequires the best Project ManagersWarning signs
Project Managers sometimes act as clowns to provide diversion.
High BS factorAppears that they just want to have their way with the
Cows.
31
Don’t throw novice project managers Don’t throw novice project managers at a bull projectat a bull project
32
Brooks’ law in action: throw a Brooks’ law in action: throw a developer at a late bull projectdeveloper at a late bull project
33
Complexity is like Amplitude and Uncertainty is like Complexity is like Amplitude and Uncertainty is like FrequencyFrequency
Bull
Cow
Colt
Dog
34
Product Innovation FlowProduct Innovation Flow
Adaptive Activities Pro
ject
Sanct
ion
RTM
CORE Activities
Idea Filter
Hot Items
ABacklog Burnup
Sales
Services
Customer Support
ProductBacklog
A Items
IterationBacklog
Flexible ScopeBacklog
New
ly D
isco
vere
d
It
em
s
Most Items for consideration in next release
B & CReleaseBacklog
B/C/D
35
Nickoliasen Model for Idea FilterNickoliasen Model for Idea Filter(Executive Summit ADC 2004)(Executive Summit ADC 2004)
Mar
ket
Dif
fere
nti
atin
g
High
Low
Mission CriticalLow High
PartnerOr convert
To Mission Critical
Focus andAllocateResources
Who Cares?AchieveParity
36
Products Tend to follow a Lifecycle pathProducts Tend to follow a Lifecycle path
Product Lifecycle
Complexity
Un
cert
ain
ty
Low
High
Dogs
Cows
Low High
Bulls
Skunks
Colts
AB
C
37
An Example Lifecycle pathAn Example Lifecycle path
Product Lifecycle
Complexity
Un
ce
rta
inty
Low
High
Low High
2003
2004
38
Portfolio Management and Dealing with DarwinPortfolio Management and Dealing with DarwinM
arke
tD
iffe
ren
tiat
ing
High
Low
Mission CriticalLow High
Invent Deploy
Offload Manage
39
Portfolio Management and Dealing with DarwinPortfolio Management and Dealing with DarwinM
arke
tD
iffe
ren
tiat
ing
High
Low
Mission CriticalLow High
Create Change
Embrace Change
Eliminate Change
Control Change
40
Portfolio Management and Dealing with DarwinPortfolio Management and Dealing with DarwinM
arke
tD
iffe
ren
tiat
ing
High
Low
Mission CriticalLow High
Ad Hoc Agile
Outsource Structured
41
Project Complexity and Uncertainty influence how Project Complexity and Uncertainty influence how to scale our agile processto scale our agile process
One size doesn’t fit all
The assessment tool is a guide
Landmark’s portfolio balance10% Bulls
10% Cows
20% Colts
60% Dogs
42
Don’t mistreat your cowsDon’t mistreat your cows
43
Don’t over stereotype:Don’t over stereotype:e.g. Not all dogs are the samee.g. Not all dogs are the same
44
Agile SteeringAgile Steering
Each iteration provides new information and takes us closer to the product vision. Uncertainty remains, and the shape of the cone is proportional.