6 I EDITOR'S NOTE 20 I TOP TEN PROJECTS IN 2006
Intere ted in innovation? Then put the What project are CIO im'e ting their Publisher
Zift Davis Media Inc.
development tool in your cu tomer' money in thi year? ccordin to a
Editors in Chief
hand, it back and let them dri e your Baseline urvey of more than 1.400 Eric Lundquist, eWEEK
company' innovation. reader, the leading three projec are John McCormick, Baseline
By eWEEK Editorial Director Eric Lundquist oice over Internet Protocol, ou ourc- Ellen Pearlman, CIO Insight
ing and data networking. Contributing Editors
SIINNOVATIONMAKES By Baseline enior Editor Todd pangier Tom Halligan
EMERGING TECHNOLOGIES PAY OFF Jack Rosenberger
A survey of nearly 400 senior IT 30 I INTO THE GREAT WIDE OPEN CopyEditor
executive investigate how emerging Migrating to an open source oftware Lisa Delgado
technologie and IT organization are tack operating on IBM equipment hojed Diredor
contributing to bu ine s innovation. aved eTrade millions of dollar and
By CIO In ight Executive Editor Allan Alter enabled the company to optimize i Andrea Mahoney
internal development proce e. Today. hodudion Manager
10 I EXPERT VOICES: eTrade enjoy ignificant performance Ivis Fundichely
GEOFFREY MOORE increa e and greater flexibilit). Art Diredor
Innovation mean very little if all your Bye WEEK enior Analy t Jason Brooks Andrew Capitos
envi ioned new product and ervice Graphic Desipers
are never actually created, ay 34 I END NOTE: INNOVATION
Geoffrey Moore. Innovation require BEGETS INNOVATION
an inve tment trategy that put your To get value from IT, focu on the All editorial content was written and produced
re ource where they matter mo t and a down tream tran formation of th by the staffs of Baseline, ao Insight and eWEEK.
people trategy that align tho e re ource bu ine that's made pos ible by No material may be reproduced in any
form without the written permission of the
with the be t kill of your employee. implementing new technology. publisher. For information about Innovations
By CIO In ight Editors Ellen Pearlman By ZiffDavi Media Editorial Director and 2006 magazine, call 212-503-5640.
and Edward Baker enior Vice Pre idem Michael Vi:ard
Zift Davis Media Inc.
28 East 28th Street
New York, N.Y.10016
4 -? INNOVATIONS2006 -? SUMMER Cover illumation by Craig 'razier
.......................................................................................... , .
I Editor's Note I Eric Lundquist, eWEEK
LET CUSTOMERS DRIVE
I OVAII N
IS THE INNOVATION WHEEL about to spin once more? applicarion deyelopment firms are numbered. In that
I think so. Most companies cite listening to their customers era, once the application was developed, it would remain
as one of their main goals. Listening and spending time in place for year or decades without change. Today's
with customers is one of the ways companies hope to innO\"arion-laced world of service-oriented architectures
deliver a product or service ahead of their competitors. is ju t tarring to hint at this change.
While listening to customers is a given, how about the What if Zillow decided that there were more precise
idea of taking yourself out of the equation? Instead of real e rate pricing mechanisms? Rather than gnash their
listening to your customers, why not give those customers teeth and hope no other company embraced the more
a basic framework, then stand back and let them design, elegant pricin tructure, zillow could easily and transpar-
market and fund your products? ently mO\"e to that pricing service. In the service-oriented
If that sounds like a radical idea, you haven't been architecture world, a new competitive offering can be
spending much time on the social networking sites embraced and become a customer feature for your company.
that have sprung up on the Internet. Com- Changing sales force automation systems,
munity sites such as MySpace.com provide a billing ystems and customer contact systems
framework upon which a community builds "ill oon become projects that can be done
informal, formal and informational networks in months, rather than in years at a cost of
about people. hundred of millions of dollars. You, as the
Recently, companies such as Zillow c tomer, will fit your application needs
(www.zillow.com) have been created around into the framework that major vendors are
the concept of putting real estate information prO\iding. And rather than locking you into a
into the consumer's hands. In this case, Zillow proprietary approach, those vendors will be
has provided a robust framework that includes trhing to show how quickly they can adapt
real estate value estimates, but also opens to your needs and requirements.
the door for buyers, sellers, banks and event real estate Does all thi ound too good to be true? It would be, if
professionals to inhabit the Zillow space. What Zillow has it weren't happening already. Are you unhappy with your
done for real estate is also taking place in online banking, Vonage digital phone service? Then try Skype. Or Yahoo.
stock markets and corporate manufacturing. Or AOL. Are you unhappy with your e-mail account on
And the innovative idea of letting the customer create Google? Then pick anyone of 10 different Web-based
the final product is not just the province of the digital e-mail system available.
world. Toyota's Scion automobile line encourages a robust Those same techniques that allow you to pick and
aftermarket of products and services for customers that choose digital phone or e-mail systems will soon be avail-
want to personalize their Scion with specialized paint able in your corporate computing environment. You, the
applications, tires and, of course, booming sound systems. customer, will be able to pick and choose applications
Toyota is trying to take the automobile aftermarket business running on virtual computing networks where all the old
from the professional few to the amateur many. hassles of incompatible systems are invisible.
Innovation in this technological spin of the wheel
ADAPTING QUICKLY TO CHANGING NEEDS doesn't mean simply listening to your customers. It means
In technology, this approach of providing the framework putting the development tools in your customers' hands
and letting the customer do the final detail work will have and letting them drive innovation. '"i'
a profound impact on the computing environment. The
days of custom work being done by highly specialized Eric Lundquist is the editorial director of eWEEK.
6 -7 INNOVATIONS 2006 -7 SUMMER
Story by CIOInsight Executive Editor Alter
...; illustration by Todd Davidson
Greg I aac i following in the foot-
tep of the mi ionarie who went
to China year ago, only he i not
there to do the Lord' work. In tead,
he explained in a phone call from
hanghai, eBay Inc. ent him there
to "evangelize to hine develop-
er ." eBay e timate that more than
430,000 people in the nited tate,
and thou and more worldwide,
make a living - full- or part-time -
elling on the online auction ite.
ow the company hope to add
tens of thou and of Chine e citizen
to that number. It want to encour-
age third-party software develop-
ers to create oftware u ing Web
service that will help high-volume
Chine e cu tomer, the ort who ell
thousand of item online, to ea ily
manage their eBa ale. bu ine partner, IT ha become a a r cent conference. "The only way
eBay al 0 want developer to fundamental ingredient in innova- we are going to pro per i to attract
create oftware that will let Chi- tion. nd today, inno ation i n't high-value inve tment and perform
ne e buyer place and track bid ju tabu ine fundamental, but high-value economic activity here
on their mobile phone . I aac ' an merican ob e ion - even a in the nited tate. nd the way
mi ion, a the director of eBay' patriotic duty. to do that i through creating n w
developer program, i "to attract "Why i innovation important? value, new products and ervices that
the developer community to build Well, quite frankly, it's the only way compete and ucceed globally."
innovative application ." we're going to have a [high] tan- This attention to innovation led
It' too early to tell if I aac 'trip dard of living and pro perity and em Insight' editors to expand the
will payoff, but few would doubt it' maintain our ecurity a well; aid annual urvey on emerging technology
worth the gamble. Whether de el- Deborah Wince- mith, pre ident of into th topic of innovation. in
op d in-hou e or obtained from the Council on Competitivene s, at previou year, our tudy look at the
R. 4 UJNnV 4. TTON" ?nn,; -4 c:.11M M 1='P
adoption of emerging technologie Why do innovative companie find for quickly implementing promising
such a oice over IP and ervice- benefit from emerging technologie new technologies. Such companies
oriented architecture, and track that elude the les inno ative? While are al 0 twice as likely to measure
whether more companie are becom- innovative companie have an IT the bu iness value of a new technol-
ing early adopter of technology. But trategy that i de igned to up port ogy after they deploy it.
thi year, our urvey of 396 enior innovation effort, that i not a cau e The e findings were echoed at a
IT executives al 0 investigated of innovation, but an effect. pre entation on innovation at the
whether emerging technologies and "Innovation tart off with bu i- May 2005 Gartner ympo ium in
IT organization are contributing to ne leader hip making a determina- San Franci co. Kathy Harri ,a
busine s innovation. tion that they have to find new way Gartner group vice president,
The re ult : Mo t IT organiza- to do thing," ays Dale Kutnick, noted that innovative organizations,
tion are actively involved in bu i- director of re earch and enior including IT organization, follow
ne proce innovation, and many vice pre ident at Gartner Inc. uch a proce model for innovation:
are contributing to product and leader, according to
ervice innovation a well. IT orga- Kutnick, al 0 reward
nization are developing sy terns employee who come
that improve or replace proces e up with innovative way "Youmust keep the learning
and are finding new way to d liver to generate revenue or going."
products and service. We al 0 ave money. - Kathy Harris, Group Vice President, Gartner
learned that more companies are That i largely
early adopter of emerging technol- confirmed by our own
ogy than in pa t urvey. urvey: While not all innovative g n rating idea, evaluating and
But the most important re ult companie provide uch reward , electing a few idea from among
wa thi : Emerging technology innovative companie are twice a them, developing and implementing
doe n't make companie innovative; likely to do so. Innovative compa- innovation, and learning from the
in tead, innovative companie make nie al 0 have a culture that up- experience so they can improve their
emerging t chnology valuable. ot port technology experimentation program over time.
only are innovative companie more and ri k-taking. "That' what make their innova-
likely to be early adopter, but IT But inno ative companie don't tion program sustainable: You mu t
executive at innovative companie depend solely on incentive and keep the learning going," Harris said.
are almo t twice a likely to report culture; they rely on proce ,too. Finally, innovative companie are
significant payoff than tho e who Companie where emerging tech- far more likely to deploy new tech-
work for companie th y do not nologie payoff are far more likely nologie . Of the 30 emerging
consider innovative. to have procedure and proce se technologie we urveyed, ju t
one - peechjvoice recognition - ha
been deployed more frequently by
)) companie that are not innovative.
sually, innovative companies are
of corporate budgets are spent specifically on innovation. at least twice as likely to deploy any
of the e technologie ; in some ca e ,
such a grid computing and elf-heal-
73% of IT departments
are involved in new product and service ing computing, the ratio i a much
a 9 to 1 or 10 to 1.
Given that innovative companie
2 * of companies are early adopters of information technology. are not only more likely to deploy
68% of companies have received significant
adopting emerging technology.
eem they are in a good po ition
to u e IT not only to remain inno-
say emerging technologies play an integral role in vative, but to maintain a u tain-
Of supporting corporate innovation. able competitive advantage in
year to come.
SUMMER ~ INNOVATIONS 2006 ~ 9
Innovation isn't just
about inventing the
new new thing, says
It requires a conscious
investment strategy, and
the will to carry it out.
r!fii]~~~~~[l!'!mL~f1Ililt!.m.rn Pearlman and Edward Baker
Ellen -+ mJ· :JE'W'~'
t!7j. Jamie Tanaka••
globalization sets in and competi- many great ideas would never have ays. MIttends to be highly bia ed
tion heats up, companies large and been commercialized had not others toward di ruptive innovation. But
mall are turning to the diligent ap- seen and recognized their true alue. it al 0 tend to be emireligious.
plication of any number of innovation Moore' new book, Dealing with almo t like the way open- ystem
practices to tay on top. Customer Darwin: How Great Companies people get religious. nd you want
innovation, innovating from the edge, Innovate at Every Phase of Their to ay at the end of the day,' o. Be
innovation road map - all are well- Evolution, ju t publi hed by Portfolio, relentle Iy pragmatic. Are you get-
meaning efforts to help companies argue that innovation by itself i not ting a return on your innovations?' M
with the innovation proce . enough. Innovation also require an
But according to Geoffrey oore, inve tment trategy that puts your
a managing director at con ultancy re ources where they count, and ForCIO- relate<b:esearc!\;ple-ase
TCG Advisor and a partner at a people trategy that align those Y.lSit--.aoiJ:i5i:glit:Co
venture capital firm MD in enlo re ources with the be t kill of all
Park, Calif., innovation mean little if your employee .
all tho wonderful new products and MThere' a kind of innovation
sel"\;ce never ee the light of day. con ulting that tend to be zealou
Con ider the notoriou fate of Xerox about how to innovate, and about
Corp.' Palo Alto Research Center: Its the freedom of imagination,MMoore
SUMMER -7 INNOVATIONS 2006-711
rRecently, Moore sat down with CIa Insight editors Ellen Pearlman and
Edward Baker to chat about why innovation really matters, and about
how best to get it done. An edited version of the conversation follows.
CIO INSIGHT: How did you choo e ate through innovation, creating a new at 1icro oft, look at Ci co y tem .
Darwinism as a metaphor for the value proposition that the cu tomer That' becau e they're not getting
economic life? prefer and your competitor don't a net new amount of competitive
MOORE: The parallels between have. That will win you new revenue advantage. They're just kind of recy-
economic y tems and ecological sys- at attractive margin . Tha~ a good cling the exi ting competitive advan-
tem are pretty well known. I wanted economic return on innovation. tage in their exi ting categorie .
to find a platform to free ourselve econd, you can innovate to neu- Look at the tech ector. In that
from the ense of entitlement in the tralize a competitor' innovation. ot ector we've alway thought of
.. economy: that we're entitled to quite a attractive. You don't nece - innovation a that di ruptive tuff
high margins, that we're entitled to arily gain an advantage, but you can that magically create new catego-
u e more of the world' re ource begin to overcome a deficit, to catch rie . The ector ha alway been
than any other nation ever has. up. a re ult, you gain more sale, about the technology adoption life-
The .. economy ha sailed with albeit not with a really competitive cycle. You tart with mainframe,
the wind for all of my adult life. And margin. But at lea t you're in the deal. then minicomputer, then the PC,
now I think, for the re t of my adult nd you can al 0 innovate in way then the L , then the Internet.
life, we're going to be ailing into that don't chang your outward com- ow it' mobility. Each change ha
the wind. What we're eeing i a sea petitivene ,though they can change had a tronger back-flu h, until
change in economic power: It's cros - your return on innovation internally. finally, when we got to the real
ing the Pacific. The U. . i not going That mean doing thing more ef- correction, the bur ting of the
to have all the economic power in ficiently, getting the ame amount of tech bubble, it wa really about the
the future. bang for Ie buck. legacy inertia of technology refu -
It' time for a new et of rule. What Unfortunately, if you look at the ing to die.
Darwini m ay i either evolve or get kind of innovation that go on in What we're eeing now i the
marginalized. Darwin doe n't care. At e tab Ii hed enterpri e , more and maturation of the tech ector. Yet
the end of the day, you perform or you more innovation i not generating the idea linger that if you're not
don't. D alingwith Darwin was my a net n w differentiating return. 0 doing di ruptive innovation, nothing
metaphor for aying tep up, or don't. you see tock price of e tabli hed of intere t i happening. And that'
enterprises being flat for year - look complete and total bunk.
SO IT'S "INNOVATE OR DIE"l
Too many e tabli hed enterpri e ay
to me, "We've 10 t the ability to inno-
vate around here." That's completely
r"It's time for a new set of rules. What Darwinism
and totally untrue. But something says is either evolve or get marginalized.
ha happened to innovation at the e Darwin doesn't care. At the end of the day,
companie that's of concern, and it
ha to do with return on innovation.
you perform or you don't."
In my view there are only three &.
kind of innovation: You can differenti-
14 -? INNOVATIONS 2006 -? SUMMER
WHAT'S THE ALTERNATIVE 1 are you going to get the re ource ? make mo t of their money through
The big problem i that too many Iy view i that you have to extract media buying, which i not a value-
companie dabble in innovation. re ource from context to repurpo e added proce . 0 er time the ad in-
They ay, "Well, we've got a little of for core. du try i going to rationalize around
thi ,and we've got a little of that." I'll give you an example. Right what' core - the creati e tuff and
But that only guarantee - ince now, 1icro oft i getting beaten up. account relation hip management.
they don't take anyone project very Everyone i aying Coogle i win- o long a media buying remain a
far - that they never e cape the ning the IQ award, and 1icro oft i fairly traightforward proce ,their
gravitational field of their ector. brain-dead. The rea on why 1icro- client will decide that it' cheaper to
They'll go down a path for a while, oft i in uch trouble is th y have to buy their media through dedicated
and then they'll think, "I like thi put all their re ource into their two media-buying firm. d agencie will
other idea, too." nd 0 they'll try a franchi e ,Window and ffice. But ay, "But that' how we make all our
little bit of that and a little bit of the Darwin ays no. You have to continue money." nd the client will ay, "Ye ,
other. nd their innovation don't to meet the revenue commitment what' your point?" Darwin doe n't
re onate. I call it morga bording. of tho e context franchi e ,but you care.
eanwhile, 95 percent of their have to do it with fewer re ource , 0 But the agencie have to charge
revenue i coming from a commodity that you can take ome re ource and for omething. I believe th y'll be-
product or ervice they're producing pend them on core, on whatever come more like con ulting firm - for
that they know damn well i never type of innovation will work for you. a while. nd then, in thi increa ing-
going to get differentiated. Thi i how they did manufacturing Iy fragmented world, a media it elf
Here' the critical di tinction: core at Ci co y tem . They centralized it, fragment more and more, tho e
versus context. I define core as that they tandardized it, they modular- client will come back to the agen-
which creates a return on innovation. In- ized it, they optimized it, and then cie and ay,"1 do want you to buy
novation for differentiation i core. And they out ourced it. the media, becau e it' gotten too
everything else i context. It may be mis- In the 21 t century, you can no damn hard. But now I need you to
sion-critical context, but it' still context. longer afford to be a ertically help me with Coogle. How hould I
The problem i that you can't integrated corporation with all of do that ad? re my click at the right
improve your economic outlook on your time, talent and management level? Do I have the right earch
the context ide. You can get more attention allocated aero s all of the algorithms?" And then the agencie
productive, but you can't change your variou function equally. Darwin will be back to buying the media, and
competitive po ition, your future, will not reward that behavior. making money that way.
without changing your competitive-
advantage equation. If you don't have WOULD THIS MODEl APPLY LOTS OF COMPANIES HAVE
any new competitive advantage com- EQUALLY TO A SERVICES COM- MISSED SUCH TRANSITIONS.
ing along, every year the environment PANY SUCH AS AN AD AGENCY? HOW CAN THEY GET BETTER AT
gets a little bit better at competing I think it applie to any indu try. SEEING WHAT'S COMINGl
again t you, and every year you get a nything can be core and anything If you allocate your re ource ba ed
little bit more marginalized. can be context. The key idea about on your revenue maker, or margin
o you've got to cut a little more core and context i that what' maker, you're driving in the rear-
co t. And you get a little more context for you could be my core. view mirror. You're optimizing for
marginalized. You cut a little bit nd my context could be your core. what ha been ucce ful. long a
more co t. nd we watch the every That a great economic relation- you're on a straight tretch it work.
powerful in titution quietly un et hip: If it' context for you and it's my But a soon a you hit a curve, it
them elve with about a one-degree core, you're going to want me to do doe n't work.
decline every year, for decade. it, and vice ver a. If you're like mo t companie , you
What' core to an ad agency? The take la t year' budget and you ay
HOW CAN COMPANIES REIG- weird thing about ad agencie right okay, let' look at la t year' budget.
NITE THEIR GROWTH ENGINESl now i their economic model i not Thu ,you've already in titutional-
You have to self-fund. But where aligned with their value model. They ized the re ource allocation of the
SUMMER ~ INNOVATIONS 2006 ~ IS
prior year. Which mean that you're
funding context before core. 0 core
get the crap after context ha been
at the trough fir t. But if you're really
doing cor vcr u context, you have to
fund core before you fund context.
o where hould you tart? You
ha e to tart with a trategy. And
then you find the really, really dirty
ecrer, which is that mo t companies
don't have a trategy. Or they have
a trategy but it's sort of an inertial
thing. But if you really are going to
do core vcr u context, and get a re-
turn on innovation, the company ha
to work from the arne trategy.
so YOU BUDGET FOR CORE
FIRST, AND THEN FOR
CONTEXT, BUT NOW THERE'S
LESS MONEY FOR CONTEXT
THAN THE YEAR BEFORE.
THE CONTEXT PEOPLE WILL
SAY, "WE HAVE TO HIT OUR
REVENUE TARGET, AND IF YOU
CUT THAT BUDGET, WE CAN'T
DO IT." THEN WHAT?
nfortunately, the correct an wer
is that the company need to find
omeone who can hit the target. "Do
you want to re ign over thi ,or do
you want to take a hot at it? Because
that' the choice."
ow, the good new about mi -
ion-critical context is that it ha
inertia on its side. If you're that big,
and you've got that much money
a ociated with you, you're an e -
tabli hed category, and establi hed
categoric have in rtia. Think about
a merry-go-round. Once you've
gOt a merry-go-round pinning, it
take Ie energy to keep it pin-
ning than it did to tart it. 0 you
ought to be able to take orne pro-
ductivity ri ks with your bu ine
without dimini hing your top-line
or bottom-line goal.
18 ~ INNOVATIONS 2006 ~ SUMMER
AROUND employee' relative intere t and pa -
Getting a return ion and kill at doing invention,
on innovation deployment and optimization. Then
requires discipline they need to detach inve tment from
in moving ideas the task and reinve t it in the kill - to
from core to
hift deployment re ource to core to
context, and then
take the next round of invention and
profits - and the deploy it.
people - back
into the next CAN COMPANIES ORGANIZE
innovation. THE TRANSITION FROM
INVENTION TO DEPLOYMENT
INVENT OFFLOAD TO CONTEXT WITHOUT LAYING
Source: Geoffrey A, Moore
D~alin9 With Darwfn, 200S OFF PEOPLE?
It i n't ea y. But the real problem is
that if you don't succe fully deploy
And the game i ,you have to. magazine. Other people put them the next invention, you're going to lay
Becau e if you don't, you can't fund out in the market, and then other off more people again in three year ,
core. optimize them. and again in three year after that,
But at too many companie , all and again in three year after that.
WHAT HAPPENS TO THE WORK- the re ource are getting tuck Here' the alternative: Take
FORCE WHEN YOU START USING in mi ion-critical context. They your top 200 leader - and thi i
CONTEXT TO FUND CORE? think they've 10 t their inventor. what Ci co doe - and ay, "Okay,
Most people cling to the task they've They have not. They've 10 t their how many inventor, how many
been ucce ful with. Eventually, that deployer . Or rather, they haven't deployer , how many optimizer
task become mi ion-critical context, 10 t them - they're tuck on mi - do I have?" nd if you move your
but they're the expert at it, 0 they ion-critical context. But they leader , you can tru t them to move
tay with it. And then it tart to get need to hift to the next genera- the re t of the re ource with them
unsetted, and they tay with it. nd tion of core. to make it happen. But if you've got
then eventually it gets out ourced. Meanwhile, optimizer are key to all your leader tuck in the wrong
nd then they're in a bad pot, getting the work out of the hand of area, you're stuck.
becau e they really have nothing el e the deployer tuck in the deploy-
they know how to do. 0 they either ment proce ,down to where it can AND THE WORKERS?
have to go with the work to the out- eventually be out ourced. That free The way we do it today i to milk an
ourcer, or they just \0 e their job. the deployer to come back to core employee's experti e and then throw
That can't be the good outcome. and begin deploying the next round away the employee. It' not fair to the
But how would you do anything dif- of invention. employee. But if you can continually
ferently? We came up with a differ- Who are the optimization people? invent, deploy and optimize, and thjnk
ent way of looking at thi : There' They are all the ix igma monk. ix about your employee a part of that
a proce model that overlay igma people hift role all the time, proce ,you can create better capital-
the evolution of any work. Every going from group to group to group, i t outcome in a way that align \vith
proce ha a ge tation or invention bringing their quality-control capabil- your workforce rather than betray it.
period, a prolonged deployment ity with them. Deployer are program That way, you're preparing them to
period, and an optimization period. manager; when they fini h one pro- urvive in chi increasingly Darwinian
II the e ta k have to happen. gram they go do another program. I busine world. :t·
lanufacturing ha all three. d- think human beings tend to gravitate
verti ing ha all three. Magazine toward one of the e three zone. Please send questions and comments on this
have all three. People in ent new Companie need to identify their article to edirors@cioinsight-:iffdavis.com.
SUMMER -7 INNOVATIONS 2006 -719
I Baseline I Top Ten IT Projects
Big money has
from Voice over IP
rStory by Baseline Senior Editor Todd Spangler ...~ illustrations by Gordon Studer ••
FOR CO RIER OF CORPORATE Tho e who identified IP telephony environment while al 0 cutting co t .
data, high on the critical-ta k list thi as "critical" to their company aid Dan liedhammer, senior infra-
year i making ure the road are they will pend an average of lLll structure engineer at Genworth,
well-paved - and that the delivery million for the year on uch project. i leading a project to extend an
truck are zooming along at peak Out ourcing project were econd, exi ting vaya phone ystem over
performance. with an average budget of $ .22 mil- IP network to two call centers, each
The Baseline Top 10 Project in lion, followed by data networking at with about 50 agents, in Phoenix and
2006 survey, which a e ed the 5.57 million. AI 0 in the Top 10 are Encino, Calif.
expected pending of information de ktop upgrade ( 3.21 million) and The project, iedhammer e ti-
technology manager, reveal that application performance management mate ,will co t 70 percent Ie than
the three large t area of funding, (2.9 million). it would to in tall two eparate,
ranked by expected pending, are Why are infra tructure project tandalone phone y tern with
related to upgrading or otherwi e a top priority? Con ider the ca e of call routing and voice mail fea-
modernizing information technology Genworth Financial. Thi year, the tures. In tead of paying 520,000
infra tructure. 10 billion in urance company will for equipment at the two ite,
o. 1 on the Ii t: Voice over u e Voice over IP electively to aug- Genworth can extend the exi ting
Internet Protocol (VoIP) project. ment it exi ting telecommunication Avaya phone witch over IP
20 ~ INNOVATIONS 2006 ~ SUMMER
For in-depth case studies, please
connection to both call center for final approval proce for hi or her witche from headquarter in
156,000. I iedhammer's team can company' information technology a hville, Tenn., ay Brad Wood,
al 0 centralize management for tho e project in 2006. The Web-ba ed enior director of enterpri e tech-
ite , in tead of requiring local tech- urvey wa conducted by Th nology operation. 10 t CCApri on
nician to maintain the voice y tern. trategy Group, a market re earch are in remote areas; a routine
Genworth, however, ha no plan firm in Englewood, Fla. admini trative ta k might require di -
to junk the 2,-00 traditional de ktop In the ection that follow, you'll patching a technician on a six-hour
phone at it headquarter in favor find example of how information round trip.
of IP-ba ed sets, becau e the exi ting technology leader are putting each "That' an exp n ive trip to add an
one work perfectly well." oice of the Top 10 Project into action. exten ion," Wood ay. a re ult,
over IP is h re, and it' a good tool," "Our facilitie would wait, and wait,
iedhammer says. "But we look at it 1 I VOICE OVER IP and wait until they needed omething
as one choice, and we're u ing it only Average spending: 11.1IM done - or until omething broke."
where it makes en e." What it is: y tern that tran mit Which leads to another ju tifica-
To be sure, Baseline's Top 10 voice over data networks. tion for the project: Many of CC 's
Projects aren't all about infra truc- Why it's hot in 2006: Companie phone witche are old and prone
ture. Bu ines -oriented application are adopting next-generation voice to outage. "The current technology
are al 0 well repr ented, with the y tern to reduce telecommunica- we ha e in the field i 15 to 20 year
other half of the Ii t taken up by tion and management co t . old: Wood ay. "It's a natural
cu tomer relation hip manage- Correction Corp. of America progr ion for us to migrate to the
ment, collaboration, upply chain (CC ), the large t operator of private new technology."
management, bu ine analytic and pri on in the nited tate, i in
compliance tracking. the third year of a four-year project z I OUTSOURCING
In December, we un'eyed 1,440 to upgrade irs phone infrastructure Average spending: .22 I
Ba eline reader , who were a cro - - which compri e ,000 phone at What it is: er ice for managing a
ection of information technology 40 company-owned facilitie - to company's information technology
executive and bu ine manager vaya oice over IP y tern . or bu ine proce function.
at companie of varying ize . ( ee key driver: to give CC ' central Why it's hot in 2006: Technology
"\ ho We urve ed" on p. 27.)Each information technology operation tandardization let bu ines e com-
re pondent wa involved in the group the ability to manage phone pare provider more ea ily.
TOP 10 PROJECTS ))
Project Average Spending"
MAJOR TECHNOLOGY INITIATIVES
BY EXPECTED SPENDING
% of Respondents
1 Voice over IP $11,112,000 $100,000 16.3%
2 Outsourcing 8,216,000 750,000 6.0
3 Data Networking 5,569,000 150,000 14.7
4 Customer Relationship
Management 5,177,000 135,000 17.3
r 5- Collaboration 4,255,000 75,000 9.5
6 Supply Chain Management 3,310,000 500,000 7.1
7 Desktop Upgrades 3,212,000 50,000 27.5
8 Application Performance
Management 2,975,000 200,000 9.0
9 Business Analytics 2,629,000 250,000 11.9
I 10 Compliance Tracking 2,479,000 160,000 6.7
'PLANNED SPENDING OF RESPONDENTS INDICATING THE TECHNOLOGY IS CRInCAl TO THEIR ORGANIZATIONS. SOURCE, BASEUNE TOP 10 PROJECTS IN 2006 SURvn.
In 2006, there' not likely to be Thi ummer, the Texa danced per econd, more than twice the
a deal bigger than General Iotor' Computing Center (T CC) expect bandwidth in the exi ting clu ter,
announcement in February that it to turn on 50 new high- peed net- chulz ay. For high-performance
would plit information technology working witche a part of a major computing, he ay, "InfiniBand
out ourcing contract - worth an upgrade to it Lone tar upercom- i looking like the way to go - it'
etiOlated 15billion over the next puting clu ter. a tandard, and a lot of people are
fi e years - among ix upplier: T C. at the niversity ofTexa moving to it:
ED, apgemini, Hewlett-Packard at u tin, provides the infra truc-
(HP) , IBM, Compuware' Covi int ture for re earcher acro the 41 CUSTOMER RELATIONSHIP
sub idiary and Wipro. country, who conduct everything MANAGEMENT
Bobby Cameron, a principal for IT from climate and weather modeling Average spending: $5.l8M
management at Forre ter Re earch, to a trophy ics imulations. "We What it is: oftware for tracking
ees GM's move as a "definitive sale and cu tomer interactions.
tipping point" in the trend of com- Why it's hot in 2006: Companies
panie oliciting out ourcing bids want to get more value out of their
u ing the tandardized Information data to better predict ale trends.
Technology Infra tructure Library 0, how's pam selling in California?
de cription ,a et of definition Thi spring Hormel Foods, the 5.4
developed by the Briti h govern- billion producer of the famouscanned
ment for information technology ham and other meats, is rollingout soft-
ervice . "That tart to allow more ware from iebel ystem, now part of
apple -to-apple compari on acro Oracl ,to more quicklyand accurately
upplier : he ay. analyze ale dara and trends.
Meanwhile, Hud on' Bay Co. Chri Boever, Hormel' director
(HBC), Canada' large t department of category ale management for
tore chain, has u ed outsourcing consumer product, ay that until
more electiv Iy: La t year, it recenrl • CRM package lacked the
entered a five-year deal with IB 1 to ability to provide "robu t" analy i
de ign and ho tan e-commerce ite of ales data. ing iebel Bu ine
selling apparel for Canada' national Analytic, Hormel will be able to link
team competing in the 2006 data in its tran actional ale y tem
Olympic Winter Game. have to upport a ridiculou range with external ource of data, uch
HB igned th deal to ell the of application: says Karl chulz, as competitive market hare figure
Olympic merchandi e in early 2005 manager of the center' high-perfor- from C iel en." ow we can bring
and needed the ite to go live by mance computing group. it all into a ingle ource," Boever ay.
ovember. "We didn't have the chulz i leading T CC's project That hould help Hormel' 500
internal capacity or the re ources to convert Lone tar from 512 tand- sales profe ional get more timely
to build the application as fast as alone erver to a 1,024-processor in ight into trend, and improve
we needed: ay interim CIO Paul blade- erver y tem from Dell, run- accuracy ince the data will be
Bellack."It would have been hugely ning Linux operating y tem . To let aggregated." e want to become
embarras ing if it had been time to ell all of tho e proce or communicate more efficient pro ider to our
Olympic gear and we weren't ready: a if they were one gigantic y tem, employee: Boever ay.
By farming out the project to IB I, he T CC need a very high-speed,
ay , the company met its deadline. low-latency network, 0 it will adopt 51 COLLABORATION
InfiniBand, a networking tandard Average spending: 4.26M
3 I DATA NETWORKING upported by IB 1, Hewlett-Packard, What it is: oftware to Jet people
Average spending: 5.57M Dell and other, to replace a proprie- hare file and other information.
What it is: y tem that tran mit tary clu ter-interconnection y tem. Why it's hot in 2006: Electronic
data among computer. The InfiniBand witche , up- collaboration i fa ter and Ie
Why it's hot in 2006: pplications plied by Top pin Communication expen ive than the alternative.
that need greater bandwidth ar (acquired by Ci co y tem last year), When the Texas- ew 1exico
driving network upgrade. can tran fer at lea t 450 million byte Power Co. wa in the proce of
SUMMER ~ INNOVATIONS 2006 ~ 25
being acquired by P 1 Re ource Getting a global view of pending i 8 1 APPLICATION PERFORMANCE
la t year, it u ed Web-ba ed collabo- even more important ince Delta filed MANAGEMENT
ration oftware to hare thou and for Chapter 11bankruptcy protection Average spending: 2.9 1
of document with out ide law firm, last eptember. part of the re truc- What it is: oftware that monitor
regulatoryagencie and con ultant . turing prace ,Delta need to pro ide application and can proactively find
Carl eider, head ofbu ine pending data to multiple con ulting potential problems.
technology er ice at the former firm. "One of the fir t things they Why it's hot in 2006: Companie
Texa - ew lexico Power, e timate want to know i , 'Who are you want more granular data to analyze
that the company hared 6,000 pending money on?'" Currey ay. y tern lowdown and outage.
document related to the acqui irion. The pend-analy i y tern will Re er eArnerica, which proce e
Printing and delivering tho e to the make that ea ier to an wer. When 3.5 million re ervation every year
20 partie involved in the deal would Delta had to pull together a com- for 150,000 camp ites and cabin
have co t at lea t $300,000, he ay. plete et of pending data in 2004, in the United tate, i u ing HP
ider' team u ed Hummingbird' it took a team of four con ultant a Open View ystems management
Enterpri e Collaboration oft- week. With the new oftware, Currey oftware to automate the monitoring
ware for the project, which co t ay ,the arne proce recently took and management of more than 200
40,000 and didn't require IT a day and a half. erver that run it Internet ite and
staffer to manage becau e it tied internal application.
directly into the company' exi ting 7 1 DESKTOP UPGRADES Thi pring, the company
Hummingbird document manage- Average spending: 3.21 expect to roll out the Open View
ment y tern. That way, for example, What it is: Computer that let Tran action nalyzer, which
the legal department et up acce worker acce corporate data. trace reque t all the way from
right it elf. "It wa ea ilya 90-day Why it's hot in 2006: Fa ter and a cu tomer' eb brow er to the
return on inve tment," eider ay . more reliable PC are replacing older pecific back-end application pro-
machine. ce e erving up the data. That
6 I SUPPLY CHAIN MANAGEMENT When Ed . Klein joined Royal will help Re erveAmerica pinpoint
Average spending: 3.31M Re ort a CIO two year ago, the problem when, for example, a
What it is: oftware to track and eight-hotel company ba ed in campground call and ay it' not
optimize with upplier.
interaction Canclln, Mexico, wa having a few getting good re ponse time from
Why it's hot in 2006: Companie problem with it 1,000 de ktop PC . the Web ite, say sy tern architect
want to contain pending by better Mo t were running indow 98; Greg Collett.
managing upplier relation . many were un table. "People were In tead of ju t indicating
Delta Air Line la t year launched 10 ing file becau e their machine wheth r a server i up or down or
a project to analyze 9 billion in were freezing up on them," Klein reporting statistics like proce s
annual pending with 6,000 up- says, noting that the humid and airy utilization, the new software will
plier in a con i tent way, u ing tropical air had corroded the com- track individual tran actions in
ho ted ervice from erticalnet. puter circuitry in orne locations. detail. For example, it could show
The sy tern, brought online last Royal Re ort replaced all of its why a pecific Web browser was
urn mer, let the airline' 115 upply- de ktop in nine month ,a project un ucces ful at ubmitting a par-
chain profe ional ea ily find out completed in the ummer of 2005, ticular tructured Query Language
exactly what Delta pent with a with mo t employee u ing the databa e request. "You can have
given vendor over a certain period. arne configuration of Dell ptiPlex it drill down and ee that thi par-
While the company' P GX280 machine with Window XP. ticular machine i running thi
accounting y tern include all that But in tead of buying them, the pecific ection of Java code lowly
data, it was laboriou to extract, ay company opted for three-year lea e and that' affecting thi ervice,"
Bob Currey, Delta' general manager for all the PC . That not only pread Collett ay. "It' very cool."
for ourcing innovation and upplier out the payments over time, but it
management. "We had to do analytical will al 0 let the company regularly 91 BUSINESS ANALYTICS
gymnastic to aggregate pending," refre h it de ktop computer. A Average spending: 2.63M
he ay. For example, contract with Klein explain: "Thi will en ure we What it is: oftware that analyze
IB 1 might be entered with different never get into a ituation where we bu ine information to detect trend
invoice codes in different countrie . have to do thi major cleanup again: or generate forecast.
26 ~ INNOVATIONS 2006 ~ SUMMER
)) BASELINE TOP TEN
PROJECTS IN 2006 SURVEY
The 1,440 Baseline readers who participated in our survey are mostly information technology managers - accounting for
two-thirds of all respondents - and 43.5 percent work for companies with more than 1,000 employees.
Title Company Size Number of Major IT Projects in 2006
Exec. VP/Sr. VPIVP 1.7%
Fewer than 500 employees 49·7%
500 to 999 6.8%
1.000 to 4.999
other IT mgmt. 27.0%
Exec. VP/Sr. VPIVP/GM 3.2%
5.000 to 9.999
More than 10.000 21.2%
Other business mgmt. 9.8%
SOUICE: BASEUNE TOP 10 PROJECTS IN 2006 SURVEY.
Why it's hot in 2006: Tool are to do their job ," ay Perroni, who access privilege to minimize" egre-
providing better integration with add that the new ver ions of AP' gation of dutie "conflicts that create
exi ting data ource . reporting tool are "phenomenally opportunitie for fraud (for example,
Halliburton, the 21 billion energy better... AP ha taken into account if an employee omehow had the
ervices and con truction company, that their product are not that ea y ability to cut him elf a check without
i planning to upgrade it AP R/3 to u e." authorization).
ver ion 4.6c enterpri e re ource La t fall, the bank installed oft-
planning oftware - which it oper- 10 I COMPLIANCE TRACKING ware from Logical App to automate
ate in more than 100 countrie - to Average spending: $2.48M check for uch conflicts, ba ed on
my AP ERP 2005, a project to be What it is: oftware that record rule about which job function
completed in 2007. change to information system, to are allowed to have certain change
In part, Halliburton decided to help show compliance with indu try privilege. The oftware al 0 records
upgrade becau e AP i pha ing or government regulation. change to the account -payable
out upport for the older ver ion. Why it's hot in 2006: Companie y tem, uch a if the amount of an
But the company al 0 expect to are looking for way to automate invoice were altered, and lets sup-
take advantage of P' Busine compliance activitie . port taffer configure the sy tem
Warehou e analytic tools that are BMO Financial Group, a Canadian but pr vent unauthorized changes.
bundled into the newer version, financial ervices company \ ho e Darlene lac Cormac, vice pre i-
ay Mike Perroni, Halliburton' ub idiarie include Chicago-ba ed dent of procurement and trategic
vice pre ident of information tech- banking firm Harri , la t year fin- ourcing at Harri , ay the oftware
nology. Currently, Halliburton ha i hed a four-year project to roll out hould cut the time to monitor con-
a 3,000-u er licen e for Bu ines Oracle Financial accounting oft- flict and y tem ecurity from two
Warehou e. ware for 20,000 orth American month to about a week.
After the upgrade, all 30,000 employee. "It's not glamorou ," h ay.
employee will be able to acces To comply with the arbane- "But it wa n't until last year that
the sy tem to analyze data in S P. Oxley ct, the company must we decided we ne ded to inve tin
"Thi will make it ea ier for people confirm that users have appropriate automating thi ." : "
SUMMER -) INNOVATIONS 2006 -) 27
I eWEEK I eTrade Financial
ETRADE FI I Laved it experience with open ource work a whole career in technology
13 million annually, realized a boo t to optimize internal development and maybe get one of tho e, 0 it wa
in performance and changed the way proces e . very gratifying to have made that
it think about development by mov- Those Ie on will be critical a bet early and to have reaped tho e
ing from a un olari infra trucrure eTrade a imilate acqui ition and benefits very early, and it ba icaJly
built up during the tOck market continue to grm to compete with ri- ju t fed our appetite for continuing
boom to an open ource tack run- al uch a meritrade. In December, down the open ource path."
ning on x 6-ba ed IB hardware. eTrade bought Browneo, the online In the late 1990 , the dot-com
Building on work it began in late trading arm of Jp lorgan ha e, for bubble wa rapidly expanding, with
2001, the financial ervice company 1.6 billion. In OctOber, eTrade clo ed eTrade right in the center of it. The
ha moved it open ource migra· the purcha e of Harri direct, another online company' trading volume and
tion up the rack and i now taking online broker, for, 700 million. On the area in which it did bu ine were
the Ie ons it ha learned through Jan. 23, eTrade reported that 2005 xpanding at a rapid pace, and 0, tOo,
revenue wa $1.7 billion, up from were its hardware requirement.
1.4 billion in 2004. et income was eTradc, like many othcr financial
For more eWEEK Labs
430 million, up from 3 0 million a ervice companie at the time, man-
analysis, please visit
www.eweek.com year earlier. aged the growth by adding capacity
[~ "It wa a win-win-win:
10 Greg Framke, of the com pan '
open ource experience.
in the form of large ener
licro y tern running
operating sy tern.
30 -7 INNOVATIONS 2006 -7 SUMMER
Lee Thomp on, eTrade's vice embrace of Linux by two industry ing ervice and the ervlet that
pre ident of architecture, likens the powerhou e : IB 1 and Hewlett- rendered the HT L - over to this
row of un ervers to the aisles in Packard (HP). With their upport, new tack," ays Thomp on. The
a large upermarket. "You could go ay Framke, moving to Linux didn't team al 0 replaced the olaris ver-
into an cTrade data center, and you mean that eTrade had to move out ion of BEA Tuxedo with the Linux
could imagine your elf looking down on it own. ver ion of the tran action manage-
the ai Ie of helve," say Thomp on "Fortunately for us, HP and IBM ment y tem, ticking with olari
in u tin, Texa . "In tead of food, came out and aid, 'We're going to for the yba e databa e portion of
you would ee un Enterpri e stand behind Linux:" ay Framke the platform.
4500 - three full rows of them." in ew York. "The econd that hap- Te t showed that the Linux box
p ned, we made our Linux deci ion, could handle about 180 u er at a
CHANGING CIRCUMSTANCES and we rolled it out very aggre ively, time, compared with 300 to 400
During the fall of 2001, a et of tarting in winter 2001." u er on one of eTrade' un 4500
circum tances prompted eTrade' IT ervers. Performance on the Linux
organization to inve tigate a move box dropped harply when there
from the un olari architecture were more than about IOu er , but,
that had been erving the company' before that point, the Linux box per-
crown-jewel application, eTrade.com, formed fa ter than the un erver.
to a mo t1yopen ource tack her the Linux box' price wa
anchored by Linux. factored in, eTrade had it elf a new
One of the dri er, ay Thomp- computing platform.
on, wa a cooling tock market "J demon trated thi on Halloween
and its effect on eTrade' bottom 2001, and it wa pretty imple," ay
line. "Trading volume were above Thomp on. "The e un 4500 were
300,000 average trade a day, and co ting about a quarter of a million
that dropped to omething in the dollars apiece, and we could run 300
range of 55,000 a day in 2001," he to 400 u er on them. And here'
ay . "Our ca h flow wa affected, thi little box that co t us about
obviou ly, quite dramatically." $4,000. It wa running fa ter, but
eTrade' engineer were aware of "Everybody was pretty floored" by the you could only run 180 people on
Linux a an emerging alternative to Linux server, says eTrade's Lee Thompson. it. Jo h Levine, the CIO at the time,
proprietary nix operating y tem . aid, 'Well, then we just need to get
" t the time, I knew that [Amazon. Fir t, however, the IT organization two, for 8,000. Do it - put a box in
com] had already moved over to had to get eTrade.com up and run- production.' 0, it wa a fairly short
Linux, and Yahoo wa on a Free- ning on Linux. meeting."
B D tack since it inception," ay Thomp on a igned an architec- With thi mandate, eTrade' archi-
Thomp on. "And 0 I knew that, at ture team from the IT organization tecture department got back to work,
ome point, architecturally, an open to tudy the fea ibility of porting putting the new tack through aeries
ource operating y tem would be of eTrade.com' central component to of te ts. In December 2001, they put
ome intere t." a Linux and open ource tack. A the Linux box into production, along-
AI 0 prompting eTrade to move part of the tudy, the team replaced ide the row of un 4500 that at
off the un erver was the addition eTrade' iPlanet (now un) Web behind eTrade's load balancer.
of ymmetric multiprocessing and a erver and application erver with "Here' this little $4,000 box, play-
32-bit me age queue to the Linux open ource equivalent running on ing, literally, with the big boy of com-
2.4.7 kernel that hipped with Red Red Hat Linux 7.2. puting and doing very, very well," say
Hat Linux 7.2, the di tribution that "We grabbed Apache, and we Thompson. "Everybody was pretty
eTrade wa te ting. Without tho e grabbed the Jakarta Tomcat ervlet floored by thi - a lot of th sy tems
feature, Linux could not erve a a y tem, and we ported a repre enta- ngineers who do our production
uitable porting target for eTrade' tive tack of our application - our operation were all going over to the
application , ays Thomp on. authentication, quote ervice , box and couldn't believe how well thi
La t but not lea twa the public product ervice, ome of our trad- little machine wa doing."
32 ~ INNOVATIONS 2006 ~ SUMMER
NECESSARY ADJUSTMENTS the other hand, if we find an open modularly, with limited commit
hen Thomp on and hi team began ource alternative where we're not right. eTrade i now charting a
porting the eTrade.com application paying for licen ing, that' the way cour e for reorganizing it internal
to Linux, they fir t copied the whole we'll go: ay Framke. eTrade i development proce e to match the
code ba e into a eparate Linux about halfway through the develop- way that open ource projects and
branch. It wa in thi branch that th ment of the Tuxedo replacement, Linux di tribution are organized.
team made the adju tment nece - which i expected to ave the com- our code ba e grow ,a the
ary for the code to run properly on pany 2 million a year. bu ine get more complex, it'll
Linux. For example, th move from become harder for u to achieve
olari to Linux al 0 involved mov- A SUCCESSFUL MIGRATION the arne producti ity and rate of
ing from the un tudio C compiler Coming off the open ource migra- change that we've enjoyed in th
to G C, or G' ompiler Collec- tion experience, Thomp on \Va in- la t few year," ay Framke. ·We
tion, the dominant Linux and open tere ted in examining the proce e want to break eTrade.com up into
ource compiler, and th team had to that produced the oftware he'd ju t eparate package, to allow differ-
sort through yntax error turned up
by the compiler chang.
During the next couple of month ,
the team ported the r t of the
eTrade.com application. In the
pring of 2002, the team brought
in 50 IB 1 x330 erver and, in
tages, hifted traffic through th ir
load balancer from the row of un
machine to the rack and a half or 0
of x86 Linux boxe .
The un erver were reallocated,
retired or old to other companie .
While co t avings wa the prima-
ry moti ation for exploring an open
ource x86 migration, eTrade found
that the move al 0 yielded ignificant
performance increa e . eTrade's leaner, meaner server platform has boosted response times.
The open ource migration al 0 af-
forded eTrade a higher level of flex- fini hed deploying 0 succe fully. ent packages to evolve at a different
ibility. Thi wa e pecially important In a te t environment, Thomp on rate of change. Thi i not how we do
becau e eTrade wa making many in tailed Gentoo Linux, a di tribu- oftware today at all - ba ically, the
acqui ition at thi time, accord- tion that expo e it u er to the whole application move forward."
ing to Framke. "Linux fit our model breakneck rate of change that open eTrade al 0 i intere ted in open
perfectly for ... caling horizontally ource oftware undergoe . ourcing some of it internally
rather than vertically: he ay. "It De pite the e change , however, developed component. Framke
gave u the flexibility to bring a lot the y tern on which Thomp on wa and Thomp on believe they have
of product online that were not ju t running Gentoo wa relativel table. omething to offer, temming from
brokerage in nature." Knowing that managing change bet- eTrade' early bet on ervice-ori-
Flu h with it initial open ource ter and fa ter than the competition ented architecture.
ucce , Trade' IT organization wa i a k y to ucce ,Thomp on and ·We've benefited in 0 many way
enthu ia tic about expanding the Framke ought to find out how open from what community proce e
model, etting it ight in 2004 on ource' methodology - it " ecret have taught u ," ays Thomp on,
the BEA Tuxedo tran action man- auce" - could be applied in eTrad ' "and I think it' ju t natural that
agement server. own development team. there might b omething that we
"We're happy to pay for oftware The ecret, eTrade learned, i that do that the open ource community
where it really help u out, but, on open ource project are organized may be intere ted in." :t·
SUMMER -7 INNOVATIONS 2006 -7 33
I EndNote I Michael Vizard
To get real value from IT, pay attention to the down tream
tran formation of the bu iness that's made po ible by implementing
TECH OLOGY I 0 no i the handmaiden of SHARI G CROSS APPLICATIONS
bu ine tran formation. If there i one thing that i But innovation doe n't top there. Really imple yn-
con i tent about information technology, it i that when dication (R ) and other variant of X IL are rapidly
orne new technology come along, we typically u e it to becoming a ne\\' lingua franca for haring data acro
do what we have previou Iy been doing - only we do it multiple application.
a linle fa ter. Rather than ha\'ing to inve tin expen ive developer
when we upgrade an application, we typically apply and complex application integration tool , people are
the arne workflow proces we u ed with the old increa ingly di covering the joy of imply sharing data
application to the new application. Then, after month acro variou application. Thi , in turn, i gi ing ri e
of truggle and fru tration, we will actually get around to all ort of new Web-ba ed information er ice, al 0
to re-engineering the bu ines proce that' a ociated known a rna hup , which are the heart of the Web 2.0
with the application, ba ed on the newfound phenomenon.
and largely unu ed capabilitie of the new In lay, CIO In ight publi hed it annual
application. urvey on emerging technologie in the
For the la t three year, a ho t of inno ative enterpri e. Not urpri ingly, the urvey found
technologies have been cro ing the proverbial that of the 4~ technologie included, virtu-
cha m. It will be intere ting to ee how much alization, ervice-oriented architecture and
downstream innovation these technologie multicore proce sors are top of mind in term
will generate. of near-term adoption, while autonomic
Consider virtualization. It's all the rage in computing and grid computing remain far-off
data centers the e day becau e it allows IT gleam in IT manager' eye.
organization to ave money on erver co t by To derive the full bu ine value of emerg-
driving up utilization rate. But downstream, irtualiza- ing information technologie ,group of executive need
tion will fundamentally change the way we think about to it down for a couple of hour and ponder the truly
u ing and managing y tern . di ruptive impact the e emerging technologie could
For example, oftware publi hers will increa ingly view have on their bu ine . In the final analy i , technology
virtualization product a platform that shield them from innovation that i not accompanied by bu ine innova-
having to directly support multiple operating sy tern . tion i a promi e only half fulfilled.
nd the nature of sy tern management will change a The ecret to getting the real value out of informa-
the y tern that needs to be managed i a higher level tion technology i not to focus olely on innovations that
of ab traction that might actually be di tributed acro lower co t , but rather to pay attention to the down tream
multiple hardware platform. tran formation of the bu ine that' made pos ible by
Beyond that, virtualization i al 0 likely to alter the way embracing new technologie .
we budget for IT hardware pending. Itimately, it may
prove to be the mi ing link that finally enable wide- Michael Vi::ard i the editorial director and senior vice
pread adoption of grid computing model. pre idenr at Ziff Davis 1edia.
34 ~ INNOVATIONS 2006 ~ SUMMER