Traditional EVM makes no sense in software (and is potentially harmful) because claiming value earned based on intermediate work products--without an assertion of quality--does not provide reasonable forecasts. Agile provides an assertable and inspectable quality. Also, by ordering in terms of highest Business Value and risk considerations along with potentially shippable increments, I believe starts to include notions of value. Still, AgileEVM measures performance against plans (that can be re-baselined every iteration if needed). AgileEVM integrates cost management. Doing it well means not giving up what Agile offers: adaptive planning, quality.

1. Bridging the Agile-­‐to-­‐PMO Communica7on Gap
Brent Barton, President, Agile Advantage, Inc.

2.  An
Agile
Story
 A
PMO
Story
 Scrum
(An
Agile
Project
Management
Framework)
 Tradi<onal
Earned
Value
Management
 Agile
Earned
Value
Management
(AgileEVM)
 Project
Case
Study
 Open
Discussion
*Slides available via slideshare.net or gettingagile.com

3.  An
Agile
Pilot
◦ One
Scrum
Team

4. This is the best,
simplest, easiest to use
applica1on we have
ever go7en in both
Customer Care and the
Retail Stores!
Whatever you all did, I
want more of that!

5. Cost
 Employee
Sa<sfac<on
Savings
 Customer
Sa<sfac<on
Revenue
New
Reten<on
Revenue
 Cost
Savings
 New
Revenue
through
efficiency
Shareholder
Value
Value
Compliance
Employee
Customer
Sa<sfac<on
Sa<sfac<on

6.  Next
Annual
Budget:
 Agile
Program
(Using
Scrum)
◦ >$75
Million
Project
◦ ~100
People
involved
◦ Quarterly
Roadmap
I am so
Happy!

7.  At
the
same
<me
as
the
“Big
Agile”
Program…
◦ Dissolved
all
PMOs
◦ Created
a
new
ePMO
◦ Moved
all
Project
Managers
to
ePMO
 Newly
Re-­‐orged
Project
Manager
innocently
turned
a
project
red
(Scope,
Schedule
issues)
◦ 9
mee<ngs
with
VP’s
and
Sr.
Directors
◦ Effec<vely
trained
this
PM
never
to
put
projects
in
red

8. Scope
Schedule Cost

9.  [Project]
Por`olio
management
is
the
coordinated
management
of
por`olio
components
to
achieve specific
organiza7onal objec7ves
 Organiza<ons
that
do
not
link
por`olio
management
to
governance
increase
the
risk
that
misaligned
or
low
priority
ini<a<ves
will
consume
cri<cal
resources
- Standard for Portfolio Management

10.  Capitalize
on
opportuni<es
 Minimize
the
impact
of
threats
 Respond
to
changes
in
the
market
 Reinforce
focus
on
cri<cal
opera<onal
ac<vi<es
- Standard for Portfolio Management

11.  Priori<za<on
to
maximize
“Business
Value”
 Effec<ve
delivery
to
minimize
costs
 Re-­‐alloca<on
of
resources
when
costs
are
too
high
or
the
benefit
is
too
low

12.  Rolling
wave
planning
is
a
form
of
progressive
elabora<on
planning
where
the
work
to
be
accomplished
in
the
near
term
is
planned
in
detail
and
future
work
is
planned
at
a
higher
level
of
the
WBS.
2008 Project Management Institute. A Guide to the Project
Management Body of Knowledge (PMBOK® Guide) —
Fourth Edition

13.  It
is
not
possible
to
completely
specify
an
interac<ve
system.
Wegner’s
Lemma,
1995
 Uncertainty
is
inherent
and
inevitable
in
sogware
development
processes
and
products.
Ziv’s
Uncertainty
Principle,
1996
 For
a
new
sogware
system
the
requirements
will
not
be
completely
known
un<l
ager
the
users
have
used
it.
Humphrey’s
Requirements
Uncertainty
Principle,
c.
1998

14.  Stack
Ranked
Priori<za<on
based
on
Business
Value
and
risk
 Self
organizing,
cross-­‐
func<onal
teams
 Defini<on
of
Done
 Poten<ally
Shippable
Increments
 Velocity
 Con<nuous
Improvement

15.  Priori<za<on
to
maximize
Business
Value
" Scrum
 Effec<ve
delivery
to
" Scrum (but no
minimize
costs
cost management)
 Re-­‐alloca<on
of
resources
when
costs
are
too
high
or
the
benefit
is
too
low

16. Total Allocated Budget!
EAC
Estimate at Complete
Management Reserve!
PMB
Performance Management Baseline
$
Planned Value
(PV)
Actual Cost
(AC) (EV)
Earned Value
Time
Time! Completion!
Now! Date!

17. Cost Performance Index (CPI=EV/AC)
CPI < 1 CPI =1 CPI > 1
Over Budget On Budget Under Budget
Schedule Performance Index (SPI=EV/PV)
SPI < 1 SPI =1 SPI > 1
Behind Schedule On Schedule Ahead of Schedule

18.  Integrates
cost
and
schedule
management
 Forecasts
in
financial
units
based
on
units
used
for
actual
cost
 Decades
of
use
 Part
of
PMBOK
(ANSI/PMI
99-­‐001-­‐2008)
 Part
of
EVMS
(ANSI/EIA-­‐748-­‐B-­‐2007)

19.  Typical
implementa<ons
expect
everything
fully
Ugh!
defined
up
front
 No
asser<on
of
quality
 Claiming
value
earned
on
intermediate
work
products
19

20.  A
planning
package
is
a
holding
account
(within
a
control
account)
for
budget
for
future
work
that
it
is
not
yet
prac<cable
to
plan
at
the
work
package
level.
The
planning
package
budget
is
<me-­‐phased
in
accordance
with
known
schedule
requirements
(due
dates)
for
resource
planning,
and
the
plans
are
refined
as
detail
requirements
become
clearer
and
the
<me
to
begin
work
draws
nearer.
A
program
may
elect
to
break
the
work
assigned
to
a
control
account
into
smaller
groupings
of
tasks,
i.e.,
mul<ple
planning
packages,
for
internal
planning
and
control
reasons.
-Earned Value Management Systems ANSI/EIA-748-B-2007

21.  There
is
no
standard
advance
planning
look-­‐ahead
period
(i.e.,
a
planning
“horizon”
or
“window”)
for
conversion
of
planning
packages
into
work
packages
that
is
appropriate
for
all
programs
or
condi<ons.
Each
organiza<on
must
determine
its
own
policies
in
this
regard.
-Earned Value Management Systems ANSI/EIA-748-B-2007

22.  Priori<za<on
to
maximize
Business
Value
" Scrum
 Effec<ve
delivery
to
" Scrum (but no
minimize
costs
cost management)
"
 Re-­‐alloca<on
of
resources
when
costs
are
too
high
EVM (cost
management but no
or
the
benefit
is
too
low
benefit management)

23. Value
Strategic
Required
to
make
good
Informs
and
Guides
Decisions
Quality Constraints
(Schedule, Cost, Scope)
Source: Jim Highsmith

24.  Mathema<cally
proven
that
Forecasts
based
on
average
velocity
(story
points)
es<mate
at
complete
EAC
(dollars)
 Key
Assump<on:
The
ra<o
of
(story
points
completed)/(total
story
points
in
a
release)
is
a
good
measure
of
Actual
Percent
Complete
Sulaiman, Barton, Blackburn “AgileEVM - earned value
management in Scrum Projects,” 2006
http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1667558

25.  Release
Baseline
◦ Budget
(BAC)
◦ Ini<al
Scope
◦ Start
Date
 Each
Itera<on
(Sprint)
◦ Points
accepted
by
Product
Owner
 meets
Defini<on
of
Done
◦ Points
add
or
removed
from
release
scope
◦ Actual
Cost

26.  Now
we
can
focus
on
Value…

28.  Integrated
Cost
and
Schedule
informa<on
provides
beser
insights
than
schedule
alone

32.  Focus
on
delivering
value
 Constraints
inform,
not
dictate
outcomes
 Quality
must
be
part
of
the
decision
process
 AgileEVM
helps
communicate
by
transla<ng
points
to
dollars

33.  President:
Agile
Advantage,
Inc.
 Former
CTO,
Development
Manager,
PMO
Manager,
Agile
Coach,
Mentor,
Cer<fied
Scrum
Trainer,
ScrumMaster,
Product
Owner
 Ac<ve
prac<<oner
delivering
value
using
Agile
and
helping
others
do
it;
brent@agileadvantage.com
www.agileadvantage.com
from
small
Product
companies
to
very
Blog: gettingagile.com
large
IT
organiza<ons
Twitter: brentbarton
" Ar<cles
• “Manage
Project
Por`olios
More
Effec<vely
by
Including
Sogware
Debt
in
the
Decision
Process”,
Cuser
Journal
2010
• “AgileEVM
–
Earned
Value
Management
in
Scrum
Projects”,
IEEE
2006
• “Implemen<ng
a
Professional
Services
Organiza<on
Using
Type
C
Scrum”,
IEEE
• “Establishing
and
Maintaining
Top
to
Bosom
Transparency
Using
the
Meta-­‐Scrum”,
AgileJournal
• “All-­‐Out
Organiza<onal
Scrum
as
an
Innova<on
Value
Chain”,
IEEE