You are on page 1of 258

agile or dead

cory ondrejka • cory.ondrejka@gmail.com • http://ondrejka.net


30,000’ view
30,000’ view

agility as key competence


30,000’ view

agility crucial to better institutions


30,000’ view

why?
30,000’ view

accelerating change
institutional incompetence
inspirations for gov2.0
the “so what”
agility is our abilities to detect,
understand and capitalize on change
better and more agile institutions detect,
understand and capitalize on change
quick choice
or
(and it’s getting harder to stay on top)
because even though we like graphs like this
overall institutional performance looks like
this
5%

3.5%

2%

0.5%
66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08
(which should scare you)
45 minutes, 234 more slides
(hope you’ve had your coffee)
some ground rules
fast and interactive
so interrupt at will
100% agreement == FAIL
/me
us navy
lockheed martin
arcade games
video games
second life
annenberg school for communication, usc
emi music
entrepreneur, board members, advisor
big, complex orgs
us navy
lockheed martin
arcade games
video games
second life
annenberg school for communication, usc
emi music
entrepreneur, board members, advisor
I mention these to establish
my non-hippy cred
before talking about
communities and collaboration
small, agile
us navy
lockheed martin
arcade games
video games
second life
annenberg school for communication, usc
emi music
entrepreneur, board members, advisor
some common elements
using technology invention, development,
cultural transformation
communication
to create products people love
/you
i will assume you rely on

mobile
email
sms
social networking
twitter
do you regularly

post to a blog
update your status
play online games
spend time in virtual worlds
know anyone who
still prints out email?
tell them the asteroid is coming
this is gov20la
30 seconds on agility
developed in response to waterfall and
other a priori development techniques
assumes you don’t and can’t know what
you want at the start of a project
so develop as little you can as a time, then
test the results with customers before
moving forward, rinse and repeat
requires a certain fearlessness
ability to experiment
comfort with (little) failures
when I talk to government agencies
about agility, I often get
“we can’t do that”
want to focus on the tools of government
data
regulation
bully pulpit
commercial excitement
going to talk about these tools in
terms of the most innovative
institution I know
wait, not that version
need to go back a few years
specifically, one Matthew Fontaine Maury

(thanks to Prof. Dr. Herbert Bukert,


University of St. Gallen,
for introduction to this history)
joined navy in 1825
spent 14 years sailing
Atlantic and Pacific Oceans
and relied on the
standard FAQ of the time
Bowditch’s Practical Navigator
he was frustrated, because it didn’t
help him answer the questions he
wanted to ask about navigation
so, he taught himself spherical
geometry and applied it to his
navigation and practical exams
with predictable results
it went over the heads of his examiners
and his naval career stalled
then things got worse
stuck ashore, frustrated with the
status quo, what was
Lieutenant Maury to do?
he started blogging
he wrote a series of articles entitled
“Scraps from the lucky bag”
where he argued for the establishment of a
Naval Academy to promote professional
training, better process, and need for
improved navigational tools
his bosses didn’t like his blogging
they exiled him to the
Navy Depot of Charts and Instruments
failing to recognize and leverage change is
not unusual institutional behavior
(we’ll come back to Maury in a moment)
gaps, especially during technological
discontinuities, are very difficult to
discover internally
are you
watching and understanding
change?
questions to think about

who will first know if the rules change:


customers, partners, or employees?

who is responsible for capturing this knowledge?

how will you collect it?


how will you act on it?
how are we handling change today?
looks ok, we see clear year-over-year
labor productivity gains

2.5%

2%

1.5%

1%
60’s 70’s 80’s 90’s 00’s

the 2009 shift index, hagel et al


so we’re good to go, right?
not even close
return on assets
5%

3.5%

2%

0.5%
66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08

the 2009 shift index, hagel et al


75% drop in RoA in 40 years
institutional incompetence
accelerating institutional incompetence
return on assets
13%

11%

top quartile bottom quartile

-15%
1965 1981 2008

the 2009 shift index, hagel et al


and lest winners feel confident,
topple rate more than doubled
only 8 of largest 25 companies in 1999
remain in top 25 in 2009
on the plus side, it appears that we do
collective recognize the problem
where “we” == “public markets”
6% annual growth in volatility
0.025

0.02

0.015

0.01

0.005
72 74 76 78 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08

the 2009 shift index, hagel et al


so where has this value gone?
5%

3.5%

2%

0.5%
66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08

the 2009 shift index, hagel et al


consumers
high value employees
radically decreasing technology costs
mean barriers to entry are vanishing
combined with a proliferation of tools
to help, it is easier than ever before to
remix, aggregate, and compete
and institutions aren’t keeping up
just like the Navy didn’t
keep up with Maury
stuck ashore, crippled, career over
except the depot
stored more than
old navigational
equipment
first tool of government: data
governments collect stuff
every naval ship was required to
submit its log at the end of a voyage
thousands
of ship’s logs
weather, winds, and currents for every day
that every US Naval vessel was at sea
so he started collating
and found patterns
and blogged his results
so, in 1844, we have a world-wide project
creating value from the aggregation of
previously valueless data
by watching, understanding, and measuring
the actions his customers were already
taking, Maury created new information that
helped them
so, let’s compare the actions of the US Navy
in 1844 with a modern example
let’s say, the record labels
revenues peaked in 2000,
generally declined since
pre 2000
post 2000
direct result of Moore’s Law
driving decreased costs and
increased connectivity

this is a new world

but change is hard to manage


CD revenues skyrocketing, new technology teams
experimenting at some major record labels
(EMI Massive Attack album delivered via stream)

Internet bubble pulls tech talent and knowledge away


(David Bowie’s “Hours” via download from EMI)

Revenues peak

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Napster
Recession begins
Revenues decline
iTunes

Guitar Hero
EMI goes DRM free

Rock Band/iPhone Games as music drivers


iTunes rich digital packages
CD revenues skyrocketing, new technology teams
experimenting at some major record labels
(EMI Massive Attack album delivered via stream)

Internet bubble pulls tech talent and knowledge away


(David Bowie’s “Hours” via download from EMI)

Revenues peak

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Napster
Recession begins
Revenues decline
iTunes

Guitar Hero
EMI goes DRM free

Rock Band/iPhone Games as music drivers


iTunes rich digital packages
CD revenues skyrocketing, new technology teams
experimenting at some major record labels
(EMI Massive Attack album delivered via stream)

Internet bubble pulls tech talent and knowledge away


(David Bowie’s “Hours” via download from EMI)

Revenues peak

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Napster
Recession begins
Revenues decline
iTunes

Guitar Hero
EMI goes DRM free

Rock Band/iPhone Games as music drivers


iTunes rich digital packages
CD revenues skyrocketing, new technology teams
experimenting at some major record labels
(EMI Massive Attack album delivered via stream)

Internet bubble pulls tech talent and knowledge away


(David Bowie’s “Hours” via download from EMI)

Revenues peak

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Napster
Recession begins
Revenues decline
iTunes

Guitar Hero
EMI goes DRM free

Rock Band/iPhone Games as music drivers


iTunes rich digital packages
what could have been learned and
built from this trend?
or telecommunications adoption rates
5,000

4,000

Mobile Fixed Line


3,000

2,000

1,000

0
1998 2000 2002 2004 2006 2008
100

Cellular Internet Mobile Broadband

75

50

25

0
1999 2001 2003 2005 2007
in the future, will we communicate by
a) cell?
b) fixed-line broadband?
c) mobile broadband?
what are users telling us?
speaking of telecom,
second tool: regulation
Maury realized how much
better the results could be
he saw great value in publishing the data
"in such a manner that each may have before
him, at a glance, the experience of all."
conveniently, John
Quincy Adams
agreed
Maury created standards for reporting
meteorological data and Adams
endowed the Naval Observatory
so collation was easier and more accurate
his data got better
a lesson here:
when at all possible,
write less code,
gather more data
quick question: who’s cheaper,
a silicon or a carbon employee?
especially if you can get data more easily
(google talks about this a lot)
let technology
can help you
bully pulpit
Maury had all of the US Navy data,
but wanted more
so he did what anyone would
do in this situation
he created a conference and
gave the keynote
where he offered his data to any
nation that submitted their ships’
logs in his standardized
meteorological format
within 5 years, nearly the entire
world was sharing their data with
the Naval Observatory vai Maury’s
meteorology API
(quite literal example of
rising tide raising all boats)
and gives a clue about what future
institutions need to look like
where are the real boundaries of your institutions?
maybe they aren’t where you think

product design sales customers


maybe they aren’t where you think?

product design sales customers

edge of
company?
are communities reviewing,
recommending, or reselling?

product design sales customers

edge of
company?
are communities remixing?

product design sales customers

edge of
company?
where could you put the line?

product design sales customers

edge of company?
if customers are acting as partners,
need to recognize it and act on it
partners require different levels of
trust, support, engagement,
information, etc
if they to succeed
benefits if you do collaborate

what can you do with fully


engaged partners?

what behavior matters?


5,000

4,000

Mobile Fixed Line


3,000

2,000

1,000

0
1998 2000 2002 2004 2006 2008
100

Cellular Internet Mobile Broadband

75

50

25

0
1999 2001 2003 2005 2007
what technologies are they
bringing into the workplace?
second life?
skype?
facebook?
twitter?
gmail?
are you trying to run a
closed desktop ecosystem?
(good luck)
and you’re missing a
chance to learn!
(think universities and walkways)
but if you really want to
live on the edge
second life
what is second life?
a user-generated virtual world
game-like technology,
but not a game
3d, persistent web
blending virtual
and real
built by a multinational,
gender-balanced audience
user generated content

almost unheard of in 2000, now well demonstrated


in SL’s case, drove economic and customer growth
produced enormously beneficial/long-lasting press
the keys to the kingdom
user generated
the unexpected
really unexpected
and more
sharing data created a network
effect and high virality, granting him access
to more and more data
it brought agility into his sytem
ensuring a cycle of data collection, release,
measurement, and improvement
all with feedback from customers
of course, more data and Maury’s
obsessive eye for detail let to
unexpected consequences
studying harpoon designs he found
that many whales in the Pacific had
been previously harpooned in the
Atlantic (and vice versa)
he saw this as strong evidence of a
navigable Northwest Passage
he was correct, of course
(he just had to wait for global
warming to open the passage in
2006 and 2008)
in 1845, not so much
which didn’t go well for Sir John Franklin,
HMS Erubus, or HMS Terror
but the search for the Northwest
Passage was only one example of new
data leading to commercial excitement
more than anyone, one group of
captains needed to go faster
the ’49-ers racing around Cape Horn to
connect New York City and San Francisco
at the start of the gold rush
New York to San Francisco
took 200 days
contests and challenges were published
focusing on the 3 month barrier
within 5 years, the time
had been cut in half
1854 the Flying Cloud
(89 days, 8 hours)
a record that stood for 136 years
Flying Cloud’s navigator was unusual
first, her name was Eleanor Cressy
second, she read Maury’s blog
she understood his data and
bet her life and Flying Cloud on it
Eleanor Cressy drove
institutional change
more challenging than
anything else I’ve talked about
there is no description of
institutional transformation that
captures how difficult it is
because it is hard to be fearless
particularly to lead fearlessly
leading across discontinuities hard
problems with latency
stars build reputation before change
sometimes, surfing a larger trend, not actually rockstars
or, the world is different, so expertise no longer applies
was this growth due to
skill or an externalities?
does prior performance
help you now?
are you bold enough to admit
you aren’t the best leader?

can you get others to


follow your example?
experienced navigators in 1854
didn’t believe they needed Maury
Cressy wasn’t limited by old thinking
Flying Cloud cut the New York to
San Francisco time in half
so how can we keep up in an
accelerating world?
we live in a world driven by Moore’s Law
Moore’s Law drives
exponential change
(which we’re bad at understanding)
we’re poorly wired for
exponential change
we predict linearly
starting from a few data points we
can’t help but linearly extrapolate
starting from a few data points we
can’t help but linearly extrapolate
starting from a few data points we
can’t help but linearly extrapolate
reality is exponential
so what we thought was a linear trend was
actually the early part of an exponential
resulting in less short-term change
but greater long-term change
leading to a fairly predictable sequence
“bob says his friends
are using facebook”
“nothing there,
fire bob”
“that zuckerberg kid was on
60 minutes. bob may have
been on to something”
“oh #$@!!”
just ask automakers
every 18 months the number of
transistors on a chip doubles
10 years of Moore’s Law == ?
120 months
= 6.667 doublings
18 months/double

6.667
2 = 101.59
2 orders of magnitude
100
times
cheaper, faster, longer lasting
100x is hard to imagine
100x mips,
same size
100x mips,
same size
100x smaller,
same perf
100x more storage is > 2 terrabytes
100x battery life is 2 years of standby,
1 continuous month of talking
also leads to radical cost reductions

100x cheaper is $2
all those computers
are connected,
mobile, and always
available, driving
even greater change
you manage change with
constant experimentation
which is hard
failure is a necessary part
of an experimental culture
you have to be ready to
fail fast
fail cheap
fail publicly
it’s all about staying out of the weeds,
having the flexibility to try new things,
and managing changing requirements
1 odds of building
= the right piece of
n software

(where “n” is estimated time in weeks)


few actions as expensive as a covered-up failure
nota bene:
“experimental culture”
does not mean
“just try stuff”
experiments need expectations,
reporting, and measured outcomes
to avoid burning a lot of money
and a willingness to challenge

“my 30 years of experience means...”


“we think next quarter will be...”
“my opinion is just as valid...”
recognition that the
plural of anecdote is not evidence
which Maury and Cressy did
so, if the us navy of the mid-19th century
can leverage viral marketing, metadata, and
government to change the world
used standards, openness, and sharing to
create increasing value for
everyone in the network
adapted to change and surmounted
institutional incompetence to apply agility
“we can’t do that” is
no longer an acceptable answer
what are you going to do?
any questions?

cory ondrejka • cory.ondrejka@gmail.com • http://ondrejka.net

You might also like