=KBOBG@;>MM>K;NLBG>LL=><BLBHGL F :R (C N G >+)*)
G>+)*)
8LI*YXYVISJ
,)%08,'%6)
K^_hkfE^`bleZmbhg3Ikh`ghlbl_hklrlm^flZgZerlml
<Zg\^kMk^Zmf^gm3KZ]bZmbhgiarlb\lZg]himbfbsZmbhg
%073-27-() Lik^Z]la^^m
:gZermb\l3
8)<8%2%0=8-'7Bglb`am_nebgm^kl^\mbhgh_]bl\biebg^l Lnko^rh_*1)
%2%0=>)8,-7Bfi^k_^\mob^ph__Zlm^oheobg`phke] ikh]n\mlmh
a^eirhnphkd
79440=',%-27%+%LZe^lHi^kZmbhglIeZggbg` [^mm^k%_Zlm^k
Zg]lfZkm^k
Experience how AIMMS
B G L B => LMH KR
enables you to apply
:g]ma^lnko^rlZrlÀ OR successfully
We recently completed the first of Analytics: 64 percent use it to keep up
what we expect will be many reader- with breaking analytics news (no small
ship surveys. As any analyst knows, feat for a bimonthly), 66 percent use
it’s nearly impossible to make im- it to learn about analytics (welcome
provements if you don’t know where to the show) and 75 percent use it to
Hassle-free deployment
things stand at the moment and you gain broad exposure to analytics top- alternatives to end-users
don’t know the problem you’re fac- ics (never too old to learn new tricks). , #
(#
, #'$
ing. The online survey, conducted in Perhaps the most telling statistic , )
March, not only gave us a good idea was how much time the respondents IZkZ`hg , !%
,
of where we are in terms of reader- said they spent reading a typical issue , %
ship satisfaction, wants and needs, it of Analytics. More than 55 percent
also provided an invaluable roadmap
regarding where we should be head-
said they spend at least 30 minutes
reading each issue and 22.1 percent
?neeIZ`^:kmphkd Fast GUI builder to visualize
results and to interact with
ing in the months and years ahead. said they spend at least an hour read- your model
We want to thank the 350-plus
readers who participated in the sur-
ing it – incredible numbers compared
to other online publications. The re- G^p:kmphkd
vey. While we’re pleased with the ini- spondents also rated the quality, val- CONOPT
tial response to Analytics and believe ue and timeliness of the content quite
we’re on track to attract a huge audi- high, and 83 percent of respondents ILOG CPLEX and many
more...
IZ`^* Seamless links to
powerful solvers
ence of analysts who are hungry for said they were likely to recommend it
our content, we realize we have plen- to a colleague, supervisor or student.
ty of room for improvement. Best of all, 70 respondents said
OBDC Point-and-click integration
The readership survey provided they were interested in contributing an OLE DB with data sources
...
some particularly gratifying results: article to Analytics, and I’m following
On a scale of 1 to 10 with 1 being “very up with them as we speak. As I say
dissatisfied” and 10 being “very satis- almost every issue, Analytics is an
Please visit our website for application
fied,” 82.7 percent of respondents rat- interactive exercise. We look forward examples, industry case studies and a
Transparent and maintainable
ed Analytics a “7” or higher and 62.9 to your comments, editorial ideas and model formulation free full trial license.
www.aimms.com
percent rated it “8” or higher. Nearly contributed articles. ❙
10 percent gave it a perfect “10.”
More insightful, from our stand- ÇI>M>KAHKG>K%>=BMHK AIMMS is a registered trademark of Paragon Decision Technology B.V.
point, was the reason readers use ahkg^k9 ebhgakmin['\hf Other brands and their products are trademarks of their respective holders.
Europe, Middle East & Africa = Haarlem, The Netherlands = Phone: +31 (23) 5511 512
Americas = Kirkland, WA, USA = Phone: +1 (425) 576 4060
Asia Pacific = Singapore = Phone: +65 9640 4182
* u : G : ER M B < L F : @ : S B G > ' < H F www.aimms.com = info@aimms.com
< H G M > G M L
DRIVING BETTER BUSINESS DECISIONS F :R ( C N G > + )*)
*)%896)7
8 K^lhnk\^l_hkLik^Z]la^^m:gZerlml
;rMahfZl:'@khllfZg
Lnko^rh_fhk^maZg*1)lik^Z]la^^mZgZermb\ikh]n\mlihbgmma^
pZrmhpZk]phkdbg`[^mm^k%_Zlm^k%lfZkm^k'
14 M^qm:gZermb\l3MphPhke]l<heeb]^
;rDZmarEZg`^
8 Bgm^kl^\mbhgh_mph]blmbg\m]bl\biebg^l`e^Zglbglb`am_khffZllbo^
Zfhngmlh_nglmkn\mnk^]]ZmZ'
20 :]oZg\^]:gZermb\l_hkLHI
;rC^k^fr?'LaZibkh
>__^\mbo^LZe^lHi^kZmbhglIeZggbg`]^\blbhg&fZdbg`\ZgÌm[^
Z\ab^o^]pbmahnmbglb`aml_khfZgZermb\l'
27 A^Zema<Zk^K^_hkfIZll^l'GhpPaZm8
Easier, More Powerful – and Far More Visual.
;r=hn`eZl:'LZfn^elhg
Ik^]b\m^]hnm\hf^l3@nZk]^]himbfblf_hk`^g^kZein[eb\Zg] From its classic Solver Parameters dialog to its modern
14 Ribbon, modeless Task Pane, and charts to visualize the
iZmb^gml4^q\^ee^gmikh`ghlbl_hka^Zema\Zk^lrlm^flZgZerlml'
6 :gZers^Mabl
Meetings Services INFORMS (Rhode Island)
12 Breakneck Hill Road, Suite 102
Lincoln, RI 02865 USA
IZ`^,
In true Excel 2007 style, select from galleries, drag and
Tel.: 401.722.2595
E-mail: meetings@informs.org the
35 <hkihkZm^IkhÕe^
ANALYTICS EDITORIAL AND ADVERTISING watch charts update instantly with new results.
consistent user interface makes learning and teaching
38 <hg_^k^g\^Ik^ob^p Lionheart Publishing Inc., 506 Roswell Street, Suite 220, Marietta, GA 30060 USA
Tel.: 770.431.0867 • Fax: 770.432.6969
President & Advertising Sales John Llewellyn Free Trials and Support for Your Eīorts at Solver.com.
39 Inssehk Llewellyn@lionhrtpub.com
Tel.: 770.431.0867, ext.209
Editor Peter R. Horner about the many other ways we can support your
horner@lionhrtpub.com teaching and learning eīorts – all at www.solver.com.
Analytics (ISSN 1938-1697) is published six times a year by Tel.: 770.587.3172
the Institute for Operations Research and the Management Senior Art Director Alan Brubaker
Sciences (INFORMS). For a free subscription, register at http:// albrubaker@lionhrtpub.com
analytics.informs.org. Address other correspondence to the Tel.: 770.431.0867, ext.218
editor, Peter Horner, horner@lionhrtpub.com. The opinions Junior Art Director Kat Wong
expressed in Analytics are those of the authors, and do not katwong@lionhrtpub.com
necessarily reflect the opinions of INFORMS, its officers, Tel.: 770.431.0867, ext.223 produce instant reports plus standard, Overlay, Trend
Lionheart Publishing Inc. or the editorial staff of Analytics. Advertising Sales Sharon Baker and Box-Whisker charts, without programming. Tel 775 831 0300 Fax 775 831 0314 info@solver.com
Analytics copyright ©2010 by the Institute for Operations sharonb@lionhrtpub.com
Research and the Management Sciences. All rights reserved. Tel.: 813.852.9942
BLMA>K>K>:EER: storage was getting cheaper and cheaper and the clear-eyed application of an appropriate living in silos, a shortage of dedicated analyti-
K>OHENMBHG;K>PBG@ “business intelligence” vendors were busy amount of mathematical elbow grease, the cally skilled resources, and managers who still
HNMMA>K>BG:G:ERMB<L8 creating data structures and reporting capa- growing piles of data could be successfully feel more comfortable relying on intuition and
BÌ=EBD>MHMABGDLH% bilities to give businesses a way to look at all mined for gold. gut feel than on structured analysis, especially
;NMFRLMKHG@L>GL> that raw data that was suddenly laying around. And then, seemingly overnight, the world when planning for the future. Many of my aca-
BLMA:MMA>K>BLLMBEE: The Internet had emerged as an amazing awakened to the potential of quantitative anal- demic colleagues also bemoan the mathemati-
EHG@P:RMH@H' platform for capturing mountains of data, an ysis to create huge business value. The past cal preparation of today’s students, suggesting
inexpensive laboratory for conducting statisti- couple of years we have seen a slew of books that the pipeline of analytical people needed
cal experiments, and a superhighway that en- such as “Super Crunchers” [2], “The Numerati” to solve this talent gap is probably not as full
abled information to be moved, analyzed and [3] and “Competing on Analytics” [4], a New as it needs to be. On the other hand, things
presented at amazing speed. And pioneers in York Times article [5] advocating how cool it is are moving so fast – new software tools, new
disparate industries – including finance, trav- to be a statistician today, and, most recently, mathematical techniques, new industry appli-
el and hospitality, gaming and supply chain huge public validation by people like Gartner cations, and new success stories – that I will
management – were demonstrating that, with and Accenture [6] of just how important this no doubt have lots to write about.
stuff called analytics is going to be. Welcome to “Analyze This!,” one man’s im-
Both INFORMS and I are scrambling to find perfect view on the rapidly evolving world of
our place in this world. In addition to launching analytics. I’m counting on you all to help keep
the online magazine that you are reading right that view expanding, so feel free to drop me a
now, INFORMS has recently engaged a major note to share your thoughts, ideas and experi-
consulting partner to help us understand how ences. See you next time. ❙
we can better serve this bigger, broader do- Vijay Mehrotra (vmehrotra@usfca.edu) is an associate professor,
main of analytics. Meanwhile, I have recently Department of Finance, Economics and Business Analytics,
School of Business and Professional Studies, University of
been hired as a professor of business analytics San Francisco. He is also an experienced analytics consultant
at the University of San Francisco, and in an and entrepreneur and an angel investor in several successful
analytics companies.
effort to keep up with what my students should
know (and what many friends in the business REFERENCES
world already know), I have agreed to write 1. See www.gartner.com/it/page.jsp?id=1210613
for the complete list.
this new column.
2. www.randomhouse.com/bantamdell/supercrunchers/
Is there really a revolution brewing out
3. http://thenumerati.net/index.cfm?catID=18
there in analytics? I’d like to think so, but my 4. www.amazon.com/Competing-Analytics-New-Science-
strong sense is that there is still a long way to Winning/dp/1422103323
5. www.nytimes.com/2009/08/06/technology/06stats.html
go. A recent Accenture study [7] of executives
6. www.sas.com/news/preleases/
in the United States and the United Kingdom AccentureSASAnalyticsGroup.html
Seemingly overnight, the world awakened to the potential of quantitative analysis to create
confirms a lot of things that many of you no 7. http://newsroom.accenture.com/article_display.cfm?
huge business value. article_id=4935
doubt have experienced: data still too often
Lik^Z]la^^m:gZerlml
optimization and simulation. They gave and algorithms that are the stuff of ana-
Joe and Jane Analyst access to analyti- lytical work. This includes “model-driven”
cal power that previously was available analytics such as optimization and simu-
only to technical experts. lation that are traditionally taught in the
Today, the ecosystem of Spreadsheet “quant course” at many North American
Fhk^maZg*1)lik^Z]la^^mZgZermb\ikh]n\ml Analytics has expanded dramatically. With business schools; “data-driven” analytics
ihbgmpZrmhpZk]phkdbg`[^mm^k%_Zlm^k%lfZkm^k' that in mind, a team of researchers at the such as statistics, data mining and busi-
University of San Francisco recently sur- ness intelligence; and functional-area
veyed hundreds of spreadsheet analytic analytics used in finance, six-sigma and
;RMAHF:L:'@KHLLF:G products in order to ascertain the current other areas of business.
state of the field. We uncovered more than Second, we’ll look at the unglamor-
SPREADSHEETS HAVE BEEN an im- by getting rid of the middleman. In the 180 commercial products from 150-plus ous before-and-after analytics activities
Model-Driven Analytics. Model-driv- shows several new spreadsheet genetic the math, products for statistical quality duplicate rows – everybody’s favorite
en analytics is for when you have a model, algorithms and other spreadsheet meta- control and even tools intended for use way to spend Saturday morning! If you
such as a forward-looking planning mod- heuristics at a variety of price points. on the factory floor. have regular business processes, you
el, and need to extract insight or decision For spreadsheet Monte Carlo simulation, Do you hate waiting in line? To con- might have to go through this monthly or
guidance. This is the playground for the the old duopoly of Oracle’s Crystal Ball and nect resources to customer waiting lines, even weekly.
algorithms beloved by operations research Palisade’s @Risk are being challenged at a wonderful spreadsheet queueing add- If you don’t have an intern to do these
professionals. We’ll start with the unsung the high and low end. Notable entrants are in offers custom functions to program exciting tasks, you can get some help
workhorse of spreadsheet sensitivity analy- Vose Softare’s ModelRisk and Frontline Sys- “queueing theory” formulas right in your from software. A half-dozen spreadsheet
sis. Our survey found several tools to “stress tems’ RiskSolver with innovative “simulation spreadsheet, plus a couple of other data management and cleaning tools
test” a model by varying inputs to see the optimization” capability. Our survey indicates academic tools. For more traditional made the cut for our survey, including
impact on a key output and presenting the that the spreadsheet decision tree space is spreadsheet operations management a couple that already come with Excel.
results in a Tornado Chart. There are tools growing with new products to compete with the survey discovered a set of spread- A half-dozen spreadsheet consolidation
and techniques for quickly computing and the business school standard TreePlan. Pal- sheet project management templates as tools, including Excel’s often-overlooked
charting the effect of one or two inputs on isade and Frontline have decision tree prod- well as an add-in for spreadsheet data consolidate tool, are also available.
an output, and running scenarios through ucts integrated with their other tools. envelopment analysis, which is a useful After you’ve sorted out the data and
your model. Business Function Analytics. Busi- technique to evaluate the performance done the analytics, you’ll need to share
Spreadsheet goal-seeking is growing ness function analytics is about tools of departments or branches based on what you’ve achieved. If you are suc-
up. The clunky Excel goal seek tool is no designed for specific functional areas of their inputs and outputs. cessful enough that your analytics are in-
longer the only option. Our survey found business. The survey found a half-dozen tegrated into routine decision-making or
new goal seeking tools that handle multi- spreadsheet finance tools for pricing vari- &)*36)%2(%*8)6%2%0=8-'7 pulse-taking processes, you’re going to
ple-goal seeks at once and can remem- ous types of derivatives, to support trad- (%8%1%2%+)1)28%2(6)4368-2+ want to regularly transfer analytic outputs
ber and repeat them. Nice! ing decisions (with integration to brokers) BEFORE YOU CAN do analytics, you need from spreadsheets into reports. If your
For the operations research crowd, and even a tool for automatic data feeds. data. It may come as a surprise to some organization is up-to-date on its buzz-
spreadsheet solvers that deliver spread- In spreadsheet marketing there is tool for readers that getting access to usable data words, you’ll want to move spreadsheet
sheet optimization (aka “linear pro- spreadsheet marketing engineering, tem- is often a big task in itself. Unless your analytic outputs from spreadsheets into
gramming”) are thriving, with powerful plates for spreadsheet marketing plans outfit has a data warehouse that actually dashboards that graphically summarize
replacements/upgrades to the free Excel and a spreadsheet advertising keyword works like the salesman promised, you’re results. Doing this by hand gets old af-
Solver. Frontline Systems has a break- pricing tool. probably going to have to clean your data ter a while, so utilize tools that automate
through product called RiskSolver Plat- If you are fan of doing things right the to get rid of junk and odd characters and posting outputs to dashboards and Web
form that handles models that made older first time, you will be interested in spread- numbers-as-text, and you may have to sites and that will open workbooks, print
solvers go belly-up. Long-time competitor sheet Six Sigma and quality tools. Our consolidate data from different sources reports and close files at the push of a
Lindo Systems continues to upgrade and survey discovered products for Six-Sigma or multiple spreadsheets into one place. button. Your days of cubicle drudgery can
improve its solid What’s Best!. Our survey greenbelts to focus on the process not And don’t forget to remove those pesky come to an end!
1%2%+-2+%2%0=8-'7 programmers to be productive and to work and then, after the spreadsheet runs invis- schools do almost nothing in this realm, and
(3-2+-8;)00%2((3-2+-86-+,8 well in organizations. This is an important issue ibly, see the outputs from the spreadsheet in too many companies the attitude is, “Here’s
MANAGERS SHOULD BE interested in the for people who spend a lot of time in spread- computations. This is the way of the future, your computer and your spreadsheet, you
productivity of their spreadsheet analysts; sheets, and it should be an important issue for and if you’re e-mailing lots of spreadsheets figure it out.” It doesn’t need to be this way
this is the important task of “spreadsheet the people who manage them. to lots of people, you need to learn about anymore. Spreadsheet programming is
development.” Managers, especially senior The most exciting news has to do with this. Our survey found that several vendors professionalizing and some companies are
managers, need to make sure that for cer- deploying spreadsheet applications. By plus Microsoft are providing variations on enjoying dramatic improvements in produc-
tain spreadsheets adequate controls are in “spreadsheet application” I mean software this service. In addition, some vendors offer tivity and maintainability. If you build large
place to ensure compliance with govern- written by one person for use by others. a “spreadsheet compiler” that locks away financial models, our survey found detailed
ment regulations such as Sarbanes-Oxley, When done naively, this inevitably leads to the code inside an EXE or DLL so the users formal development methodologies and you
as well as sound risk management for fi- headaches and confusion. A terrific solu- can use the spreadsheet but can’t see or might want to invest in learning one of them
nancial decisions. tion is to upload your spreadsheet app to a change the formulas. or at least take a look at Jonathan Swan’s
Spreadsheet Development. Spreadsheet server and provide a simple interface. Users The other big news has to do with spread- “Practical Financial Modelling” (Swan,
development is about supporting spreadsheet anywhere in the world can enter their data sheet development practices. Business 2008). For the general business modeler,
there are some terrific books about how Courtesy of the Sarbanes-Oxley legisla-
to approach the modeling/thinking as well tion, the U.S. government requires that
as the spreadsheet engineering, including anything in the financial reporting loop has
“Spreadsheet Modeling Best Practices” to be locked down. This means no spread-
(Read and Batson, 1999) and “Modeling sheets or only spreadsheets with tight
for Insight” (Powell and Batt, 2008). controls. Risk managers want to make
According to the survey there are sev- sure that spreadsheets used for high-val-
eral tools, some very sophisticated, for ue or high-risk decisions (e.g., pricing of
spreadsheet checking and auditing. These financial derivatives that lead to million- or
tools identify anomalies in the formulas billion-dollar bets) are bulletproof.
and flag things that you should be con- This requires a certain amount of man-
cerned about. Some of the more powerful agerial commitment and process skills,
tools provide model maps, summary data and no tool is going to help with that. It
and stuff you never knew you needed until also requires getting an inventory of the
you saw it. One of these should be in the spreadsheets in your organization and de-
bat utility belt of every spreadsheet super- termining which need managerial attention
hero. The survey turned up handy “diff” or control and which don’t. That can be a
tools for spreadsheet comparison that tell challenge, as some servers have as many
you what’s changed in two versions of the as 15 million (that’s not a typo: 15,000,000)
same spreadsheet; like yellow sticky notes, spreadsheets. This requires a tool that
these are indispensable when you need looks at everything on a server or at every
one. The survey also found a grab bag of computer on a network and gets a snap-
spreadsheet productivity tools that claim to shot of all the spreadsheets, as well as the
help you work smarter and faster. spreadsheets that sneaky users may have
Control & Compliance. For someone tried to disguise as something else. Our
keen on analytics, control and compli- survey found powerful tools for spread-
ance is not obviously exciting. But there sheet discovery and management that
are important things going on in this area. help with this. They’re not cheap, but they
sure are impressive. They can also provide
Ln[l\kb[^mh:gZermb\l things like access control, change logs, ver-
sion control and other capabilities that are
It’s fast, it’s easy and it’s FREE!
routinely provided for professional software
Just visit: http://analytics.informs.org/
developers. But then again, if you’re using
spreadsheets to run your company, shouldn’t warehousing systems. It is an open ques- the open source OpenOffice.org products REFERENCES
you treat and support (and manage and con- tion when CIOs, CFOs and other senior might yet take off. 1. Conway, D. G. and Ragsdale, C. T., 1997, “Modeling
optimization problems in the unstructured world of
trol) your spreadsheet analysts as the profes- managers recognize the amount of highly For more details on the categories, spreadsheets,” Omega, Vol. 25. No. 3, pp. 313-322.
sional software developers that they are? paid time that is invested in spreadsheet products, information sources and “beyond 2. Frontline Systems, 2009, “Optimization Problem
Types Overview,” www.solver.com/probtype.htm,
For medical or pharmaceutical re- model assets. If and when they do, expect the spreadsheet,” visit www.usfca.edu/bps/ accessed April 1, 2010.
search, there’s another angle. The FDA a surge of interest in development process- spreadsheet-analytics. ❙ 3. Powell, S. and Baker, K., 2009, “Management
Science: The Art of Modeling with Spreadsheets,” 2nd
has strict requirements on data integrity. es and productivity. Thomas A. Grossman (tagrossman@usfca.edu) is an edition, John Wiley & Sons.
Spreadsheets can still be used in medical We now have good tools for discov- associate professor of Business Analytics at the University 4. Powell, S. and R. Batt, 2008, “Modeling for Insight: A
of San Francisco. He teaches courses in Spreadsheet Master Class for Business Analysts,” John Wiley & Sons.
research, but they need to be locked up ering all the spreadsheets on a server. It Analytics, Business Analytics and Spreadsheet Engineering & 5. Read, N. and J. Batson, 1999, “Spreadsheet
inside a virtual vault. Products that can should not be too long before there are Management. He is developing a new MBA major in Business Modelling Best Practice,” www.eusprig.org/smbp.pdf,
accessed March 29, 2010.
Analytics. He does research on how spreadsheets are used
do this for you are available. good tools for mapping spreadsheet in- in organizations and how they can be used better. Web site: 6. Swan, J., 2008, “Practical Financial Modelling, Second
formation systems as a set of interlinked http://web.usfca.edu/bps/faculty/Thomas_Grossman/ Edition: A Guide to Current Practice,” CIMA Publishing.
MphPhke]l<heeb]^ ;,%8-78)<8%2%0=8-'7#
AN ESTIMATED 75 percent or more of an
organization’s data is unstructured (im-
the source of the data (textual, numeri-
cal, categorical, etc).
With text analytics, we are able to ex-
tract meaning out of large quantities of
ages, content of Web documents, stan- various types of textual information. We
Bgm^kl^\mbhgh_mph]blmbg\m]bl\biebg^l`e^Zgl
dard documents, audio, video, emails, combine it with structured data within an
bglb`am_khffZllbo^Zfhngmlh_nglmkn\mnk^]]ZmZ' call center or claims notes, etc.). The vol- automated and repeatable process to en-
ume at which this type of data is growing hance business insight.
;RD:MARE:G@> is alarming. We are constantly barraged
with digital content in the form of e-mail, 8)<8%2%0=8-'7%((6)77)7
news articles, contracts, privacy notices, &97-2)77',%00)2+)7
“TEXT ANALYTICS” could be the other. To simplify, I’ll use the classic marketing offers, tweets, blog posts and TEXT ANALYTICS addresses two major
contained within them is more easily acces- Prior to more advanced analysis or au- as “unsuccessful attempts to contact concerned that the driver “left” the win-
sible by those who need it. Taking the lead tomated categorization, the text will likely customer.” dow open. The order of the words in the
from lessons learned on the Internet, the need to be preprocessed to address data Categorizing the content of docu- sentence may also be significant. For an
key is associating words with documents quality issues and standardization so the ments effectively relies heavily on the insurance claim, it is extremely important
that provide information about the content analysis will produce more accurate re- ability to understand the meaning of the to know if “Driver 1 hit Driver 2” or “Driver
that resides within them. Better search with- sults. As in traditional types of analysis, words in text, individually as well as in 2 hit Driver 1.” Notice that the words are
in the enterprise can also be enabled with up to 80 percent of the time can be spent context. Each language has its own nu- identical, but the order makes a huge dif-
better strategies for tagging the content. in the data preparation phase of a proj- ances, and within those languages jargon ference for the insurance company in as-
Tools for automatically categorizing content ect. In addition to correcting misspelled exists that may be unique to companies, sessing fault for the accident.
that create and maintain rich metadata as- words, much of the data preparation in- to geographic areas, or among functional Categorizing documents from informa-
sociated with the documents are becoming volves standardization or transformation areas across companies. Who among us tion contained within them can be achieved
a central piece of a successful enterprise to a consistent set of terms or identifying has not had to ask co-workers or busi- through a combination of statistical models
search deployment. specific ideas or concepts. For instance, ness associates to explain an unknown and business rules. As with traditional mod-
you might want to standardize “LOL” to acronym or saying? How should one in- el development, sample documents are
8,)463')773*%2%0=>-2+8)<8 “laugh out loud,” or expand abbreviations terpret “blue screen of death”? Should it examined to train the models. Additional
WHILE COLLECTING and searching for in- in patients’ medical files from “upr” to “up- be associated with a computer “crash” or documents are then processed to validate
formation is very important, it’s only the dated patient record” or “lvm” to “left voice “hang”? The human resources depart- the accuracy and precision of the model,
first step in gaining business value. mail” for better consistency in analysis. ment might talk about layoffs as “down- and finally new documents are evaluated
The process of performing analysis on The contact center may be conducting sizing,” “right-sizing,” “employee turnover” using the final model (scored). Models can
text to discover insights turns out to be sim- a loss-prevention campaign to custom- or “planned attrition.” In order to analyze then be put into production for automated
ilar to analyzing traditional data types. First, ers who are delinquent on payments. It information about layoffs that resides in processing of new documents as they ar-
you need to explore the documents. This may be interested in tracking how many text, HR needs to associate all terms that rive. The models can be monitored for con-
might be in the form of simple word counts attempts were made to contact each cus- might be used to refer to layoffs. Typical- tinued performance.
in a document collection or manually cre- tomer. The contact center would want to ly, subject matter experts are required to
ating topic areas to categorize documents be able to associate “lvm” and “left voice identify and interpret the unique terminol- (-7'3:)6=
by reading a sample of them. For example, mail” both as unsuccessful attempts to ogy from a particular domain. SOME BUSINESSES may have a predeter-
what are the major types of issues that contact. It might also want to classify Language can be ambiguous. Words mined categorization structure that they
have been identified in recent automobile “no answer,” “bad number” and “phone in and of themselves may have different use to organize their documents. For in-
warranty claims (brake failure, air bag de- no longer in service” with unsuccess- meanings when used in the context of a stance, the technical support center might
ployment, engine failure, sudden accelera- ful attempts. In this manner, synonym sentence. Some words can be used as have a list of the top 10 product defects
tion)? As part of the exploration effort, you lists or business rules would be gener- different parts of speech. In a warranty identified through customer interactions.
will likely find misspelled or abbreviated ated to match against text files to identify claim, you may be interested in defects in Additional or emerging trends could be
words, acronyms or slang terms. various results that would be categorized the “left” wheel of the automobile but not identified through the use of text mining to
services is the first step in creating a di- but “long” boot time would be a nega- near future, rather than only the subject collided, and clever decision-makers
alogue with that audience by listening, tive attribute. of research papers. will begin to see the biggest bang for
learning and then engaging with them. Early projects with sentiment analy- Today text analytics provides forward- their buck. ❙
This dialogue can better inform targeted sis may start by categorizing positive thinking organizations with a framework Kathy Lange (Kathy.Lange@sas.com) is a senior director
marketing initiatives to customers and and negative dimensions of the feed- to maximize the value of information in SAS’ Business Analytics Practice. She has more than
25 years of experience selling and implementing analytics
prospects, enabling the organization back and counting how many responses within large quantities of text. This tech- solutions. The SAS Business Analytics practice assists
to communicate at a significantly lower were positive or negative with regard to nology helps automate the process by customers in defining their business problems and crafting
strategies for solving those problems with integrated
cost with increased speed and effec- a particular product, feature or service. extracting relevant information and in- SAS solutions that include business intelligence, data
tiveness than traditional marketing. It More advanced projects may track the terpreting, mining and structuring infor- integration and advanced analytics.
can also enable a more rapid response sentiment over time to track trends and mation to improve findability and reveal Copyright © 2010, SAS Institute Inc. All rights reserved.
to perceived customer issues and com- patterns or compare the sentiment of a patterns, sentiments and relationships A version of this article appeared in BeyeNetwork.
Reprinted with permission.
petitive threats. company’s product with the sentiment among documents. Two worlds have
of the competitor’s product.
Sentiment analysis builds on both Public relations firms use sentiment
human-defined business rules and analysis to track positive and negative
computer-generated statistical models mentions of their company, company of-
to automatically identify positive and ficials or products in the media or other
negative sentiments expressed in free- publications (pharmaceutical brands,
form text. Keywords are identified in for instance). They try to determine if
the context of the text to classify posi- their marketing messages are being in-
tive and negative concepts. Text might terpreted positively or negatively and
explicitly contain words that identify the what effect that has on their brand in
mood or sentiment, for instance, “mad” the marketplace.
or “angry.” You may also be able to infer
attitudes and emotions based on action '32'097-32
words or phrases such as “screamed” BUSINESS LEADERS COMMITTED to fact-
or “hung up.” Some phrases might be based decision-making are recognizing
more subtle, like “not impressed by” the power hidden in text to yield insight
or “answers were unsatisfactory.” And into marketing, customer service, public
some words can be ambiguous until as- relations, product innovation and compe-
sociated with a product feature or ser- tition. Techniques for analyzing voice and
vice. In the context of a computer, “long” video and other unstructured content will
battery life would be a positive attribute, be more commercially available in the
*
/
!
*
'+-
2'-
!
"
*
.
0
3
1
4
)"5
6
768
6)"5)9:A;9<=
"
;
>
9
.
?@
&'
'(
.
)
*
#&+
#####
###
###
#####
#
#
# #
# # # # # # # # # !# # #
"# #
#
L N I I ER < A : B G = > F : G = F : G : @ > F > G M
:]oZg\^]:gZermb\l
_hkLZe^l
Hi^kZmbhglIeZggbg`
>__^\mbo^LHI]^\blbhg&fZdbg`\ZgÌm[^
Z\ab^o^]pbmahnmbglb`aml_khfZgZermb\l'
SALES & OPERATIONS PLANNING Despite the many S&OP system im- and important. Companies also need to a specific application defines the S&OP
cussed in [4], both these criteria can be the markets. The company’s distribution .,%)6/.( ( -66
3 ,
2-6
6)$0 ,-$.3 # 6)$0 ,-$.36*!6 2-6.6
/-.$)
met by advanced analytics that are im- centers (DCs) are organized in three lev- "/.( .(/ / %3 , (*(-/. 2- /
bedded in a business intelligence sys- els: plant DCs, regional DCs, and local &%$(%'-(( %$(-
' $)') +-(( %$(-
%.
'(
0$'$)6*+*0 , )6($'*1$.4
tem to: (1) facilitate efficient extraction, DCs. In addition, the company has ar- 2-6.. 6)$0 ,-$.3 *,.#1 -. ,)6)$0 ,-$.3
$+ .2-.. / &-($'*1$.4 )*,.#1 -. ,) /
transformation and loading of inputs rangements with third-party warehouses
$'65 .$)&3
from ERP systems to the S&OP deci- to handle surge inventories for peak de-
%$)' *)-(( %$(-
'
2-6
6)$0 ,-$.3
/(,6/.#/,()
-$' .(/ /
sion database; and (2) report metrics mand. These DCs are designated with # 6)$0 ,-$.36*!6 2-6.6
/-.$)
&/(, /-.$)/. 2- /
'-'%'#- (%$-
'
and other results from the descriptive an S in Figure 2. 6,$6$& '
*)%' "(-
'
and SCNO models in easy to use dash- The optimization model was construct- # 6)$0 ,-$.36*!6 2-6.6
/-.$)
*#)6- ) $)
$& ' ($'/. 2- /
# 6)$0 ,-$.36*!6 2-6.6
/-.$)
boards with drill down capabilities. The ed by merging 12 monthly sub-models %#- ($'/. 2- /
*$' ( $-
'
merger of business intelligence systems based on the network in Figure 2 plus - (%$
,#)6/.)*"'/
# 6)$0 ,-$.36*!6 2-6.6
/-.$)
inventory balance equations linking con- /'6
6 )- )
,#)& ($'/. 2- /
# 6)$0 ,-$.36*!6 2-6.6
/-.$)
A^eiIkhfhm^:gZermb\l secutive months. Forecasted market de- +% )- ) ($'/. 2- /
mand was backed up to demand profiles
It’s fast and it’s easy! Visit:
for the local DCs; hence the dotted lines
http://analytics.informs.org/button.html
in the figure. This construction assumed
The total cost minimization approach for customized versions of their products.
is easier because it is simpler and directly The company must decide how much to
extends current and past modeling prac- charge for customized products, where the
tices regarding supply chain management. products will be made and when the prod-
Demand management may be handled by ucts can be delivered. Typically, the com-
manipulation of inventory balance equa- pany faces a portfolio of such requests and
tions for each product (or product family) should employ integrated decision-making
in the SCNO model. For each product and methods that simultaneously evaluate all
at each location where the product may be such requests received over a meaningful
held, this familiar equation is period of time. By “discretizing” the pos-
Ending inventory = beginning invento- sible selling price range for each custom-
ry + product received or made – product ized order, zero-one decision variables in
shipped or sold an SCNO model can select if, where and
when each order might be manufactured
Cost minimization analysis may implic- and at what net profit.
Figure 2: S&OP network of a large beer manufacturer. itly evaluate net profits by flexibility built into Consumer package goods companies
the model regarding if and when certain in- are candidates for S&OP models that in-
that the customers and markets to be ef- '3781-2-1->%8-32:7 termediate or finished products are shipped tegrate supply chain plans with marketing
ficiently served by each DC were known. 463*-81%<-1->%8-32 or by running multiple scenarios with vary- and sales plans for the coming months.
The resulting LP model had approximate- AT THIS STAGE IN THE DEVELOPMENT of advanced ing demand profiles for finished products. Market response models that predict mar-
ly 50,000 rows and 125,000 columns and analytics for S&OP, an unresolved issue when The net profit maximization approach ket share and market volume as functions
could be generated and optimized in just a constructing an SCNO model is: Should a provides more explicit modeling of integrat- of product prices, advertising, promotion
few minutes on a high-end PC. company use an objective function that ed supply chain and demand management and sales force effort can be linked to sup-
Starting early in each calendar year, the • minimizes the total cost over the decisions and therefore provides more com- ply chain models that determine the costs
S&OP model was run every month using planning horizon of meeting sales prehensive S&OP analysis. It is more chal- of supplying the demand generated by the
updated data about inventories, forecasted which are treated as fixed and given, lenging to construct these models and to marketing plans. The objective of the inte-
demands, manufacturing capacities and or use them once they have been validated. grated optimization model is to
other parameters. The objective function • maximizes the net profit over the It is important to recognize that demand maximize net profits = gross profits
was to minimize the total supply costs of planning horizon from sales – some of management modeling depends heavily from product sales
meeting forecasted demand. It provided which are treated as discretionary and on the industry. For example, companies – marketing and sales costs – supply
S&OP solutions that significantly lowered may incur avoidable marketing, sales, that manufacture commodities such as chain costs
these costs compared to earlier years be- product customization or non-standard chemicals or building materials often face See [2] for details about market re-
fore the advanced analytics were used. service costs ? decisions about requests from customers sponse models; see [3] for mathematical
programming methods that integrate Master data files list the names
market response models with supply of the company’s suppliers, facilities
chain optimization models. (plants and DCs), and markets and their
geographical coordinates. They list the
8,)7
34()'-7-32(%8%&%7) names of the raw materials, work-in-
ALTHOUGH THE FORM and content of process and finished products that flow
the S&OP decision database will de- through the network, along with their
pend on the nature of the company’s units of measure. They also list the
business and on the software they ac- names of: the processes at each facil-
quire or implement to support decision- ity; the bills-of-materials used there to
making, some general principles have transform products; and resources that
emerged as the result of successful are consumed in carrying out these
SCNO applications. The input data transformations. Geographical connec-
for the SCNO model is typically parti- tions are provided by lists of the links
tioned into master data files describing in the in-bound (suppliers to facilities),
the structure of the company’s supply inter-facility (facilities to facilities), and
chain and scenario data files contain- out-bound networks (facilities to mar-
ing numerical values for relevant costs, kets). Finally, the master data files list
activities, resource constraints and fi- other structural data such as available
nal product demands. The master data transportation modes and the number
files are permanent or semi-permanent of time periods in the SCNO model.
with changes made infrequently. By The scenario data files provide nu-
contrast, multiple scenario data files merical parameter values for each pe-
are created and analyzed as part of riod in the planning horizon including:
each month’s planning process. Out- • raw material capacities and costs for
puts from SCNO model optimization of each supplier,
each scenario are stored in extensions • facility indirect costs,
of the scenario data files. • process, product transformation,
and resource capacities and costs at
Ln[l\kb[^mh:gZermb\l each facility,
• inventory throughput and safety-
It’s fast, it’s easy and it’s FREE!
stock costs and capacities, and
Just visit: http://analytics.informs.org/
• transportation costs and capacities.
For multi-national networks, the costs For this purpose, the company needs to impose decisions if and when the dis- ,-)6%6',-'%040%22-2+
may be specified in multiple currencies to create an S&OP Coordination Com- cussion becomes prolonged and counter- ;-8,%(:%2')(%2%0=8-'7
with a currency exchange table for each mittee and an S&OP Analysis Team. A productive. One can expect, or at least HIERARCHICAL PLANNING REFERS TO the
period. The scenario data may also include typical S&OP Coordination Committee hope, that contentious S&OP problem identification and application of consis-
managerial constraints on the shape of the will be comprised of the VP for Sup- solving will occur infrequently as the re- tent and effective strategic, tactical and
S&OP solution such as the number of plants ply Chain Management, key managers sult of monthly planning exercises that operational plans. The discussion thus
where a product may be made. Finally, nu- from procurement, manufacturing, dis- look out many months in anticipating far has addressed the use of advanced
merical data describing demand forecasts tribution, marketing and sales, plus the problems and finding effective resolu- analytics in a company to support tactical
by product and time period are also speci- head of the S&OP Analysis Team. The tions. Conflicts in team decision-making S&OP over the coming months. These
fied. These data may be expanded to incor- S&OP Analysis Team will be comprised can also be reduced by creating and ap- analytics can be transformed into and/
porate net profit maximization options such of knowledge engineers who know the plying managerial incentive plans that or linked to similar analytics that address
as product pricing. company’s business and, depending on promote integrated S&OP. strategic and operational planning.
the nature of that business, have rele-
(-71%280-2+ vant, in-depth knowledge about analyt-
()'-7-321%/-2+7-037 ics for demand forecasting, data mining,
MANAGERIAL DECISION-MAKING in Figure 1 supply chain management, managerial
refers to integrated decision-making among accounting, mathematical programming
supply chain and marketing and sales man- and marketing science.
agers. Advanced analytics and supporting Each month, the S&OP Analysis Team
processes serve to dismantle decision- updates the S&OP decision database
making silos that are still present in many and, working with the S&OP Coordination
companies. The SCNO model provides the Committee, optimizes multiple scenarios
inspiration, or the abstract representation, describing supply and demand manage-
for analyzing the company’s S&OP prob- ment plans for the coming months. These
lems in an integrated manner. The perspi- scenarios are created in response to
ration, or hard work, for implementing the feedback from S&OP managers to the
advanced analytics is creating, validating Coordination Committee regarding the
and updating the decision database that desirability and feasibility of plans sug-
feeds the SCNO model. gested by earlier scenarios. Although
Additional effort is needed to devel- discussion about each month’s revised
op and maintain S&OP processes that S&OP solutions for the coming months
exploit insights provided by advanced should be encouraged, the Coordina-
analytics about integrated decisions. tion Committee must be given authority
Tactical S&OP analysis will often model can set goals and constraints
reveal the need to strategically re-de- for these models over the coming
sign the company’s supply chain net- days and weeks that reflect efficient
work over the coming years. Related tactical plans for the coming months.
decisions regarding the company’s
portfolio of products or the acquisition '%7)789(=4)6(-+%3
of a complementary company may PERDIGAO, A BRAZILIAN FOOD manufac-
also require such analysis. A tacti- turer with more than $6 billion in yearly
cal SCNO model can be expanded to revenue, produces poultry, pork, beef
address strategic design options by and processed dairy products. Its prod-
re-defining the time periods and asso- ucts are sold in Brazil and in 110 coun-
ciated data in the model and expand- tries around the world.
ing the decision database to describe As discussed in [5], starting in 2003,
the new strategic options. Converse- Perdigao began developing S&OP pro-
ly, a strategic network design model cesses to deal with many supply chain
can be readily converted to a tactical challenges including:
S&OP model by fixing the design de- • long lead times up to 18 months for
cisions and re-defining the time peri- acquiring raw materials,
ods and associated data. • tight constraints on slaughtering and
When moving from tactical to op- freezing capacity, To find an expert to help you,
erational analysis, the challenge is • shelf-life constraints on finished
to translate aggregate and holistic products,
log onto INFORMS O.R. Professional Database at
tactical plans to detailed operational • no formal sales forecasting process http://www.scienceofbetter.org/find.
ones. Detailed models for inventory or analytics, and
control, production scheduling, per- • a push rather than a demand-driven
sonnel scheduling, vehicle routing, system for replenishing distribution
etc. are available for examining oper- centers. INFORMS is the foremost association
ational plans. An S&OP optimization Perdigao’s installed SAP system of O.R. and analytics experts. Our members
contained the transactional data needed
to generate and apply advanced ana-
literally wrote the book on how analytics and
A^eiIkhfhm^:gZermb\l
lytics. The first step was the design and the principles of operations research are used
It’s fast and it’s easy! Visit:
http://analytics.informs.org/button.html
implementation of an S&OP process to improve corporate or governmental
spanning sales, marketing, logistics,
decision making.
+. u : G : ER M B < L F : @ : S B G > ' < H F
http://www.informs.org
L : E > L H I > K :M B H G L I E : G G B G @
production, finance and procurement for constraints appropriate to short-term and SKU level facilitated the computation of participants in Perdigao’s new S&OP
both international and domestic markets. tactical planning. Forecasting models implementable short-term operational process employing advanced analytics
The goal of the new process was to ex- were also used to predict market prices plans as well as analysis of intermedi- gained widespread user adoption. Repre-
tract insights from the sales organization for the company’s products. The S&OP ate tactical plans. However, the result- sentative results [5] include a 40 percent
that could later be used in developing transformation team then focused on dis- ing model size inhibited the running of increase in inventory turns in the com-
analytics for demand forecasting. It also tribution planning processes that would multiple scenarios each month. Plan- pany’s domestic markets, a 4 percent
promoted communication and a greater change supply chain planning to one that ning continuity was reflected in the main decrease in stockouts, and a 31 percent
sense of coordination between the sales was demand-driven from the previous scenario run each month by consensus decrease in shelf-life losses. ❙
organization and those in production, lo- one that was push-driven. among the S&OP managers in meetings Jeremy F. Shapiro (jfsinc@attglobal.net) is a professor
gistics and purchasing. After three years transforming their held prior to the run. emeritus in the Sloan School of Management at MIT, a
former co-director of MIT’s Operations Research Center
The next step was to leverage Perdi- S&OP process, Perdigao took a large next Following successful data validation and author of the textbook, “Modeling the Supply Chain.”
gao’s ERP system to develop demand step in implementing a Supply Network exercises and extensive user training,
forecasts and create a decision data- Planning Optimizer to address both short-
base with timely and accurate data de- term (three months) and long-term (18
scribing sales orders, distribution center months) issues involving supply chain and
inventories, manufacturing and logistics demand management decision-making
across Brazil. The Optimizer was a large-
REFERENCES scale LP model that addressed three sup-
1. Barrett, J. and S. Steutemann, 2010, “Conquering pliers groups, 16 plants, 20 DCs, seven
the Seven Deadly Challenges of Sales and Operations
Planning,” AMR Research, January. Document available markets, and two exporting ports. The ob- WHERE IS YOUR CAREER LEADING YOU?
at www.amrresearch.com.
jective function of this model was to maxi- http://www.informs.org/continuinged
2. Lillen, G. and A. Rangaswamy, 2004, “Marketing
Engineering,” Trafford Publishing. mize profit with constraints on product mix
3. Shapiro, J., 2007, “Modeling the Supply Chain,” 2nd to ensure long-term stability of total product
edition, Duxbury Press.
4. Shapiro, J., 2009, “Business Intelligence & Modeling
offerings. Constraints capturing shelf-life
Systems Synergy,” Analytics, fall, pp. 18-23. and inventory balances were also incorpo-
5. Tohamy, N., 2009, “Perdigao Builds a
Demand-Driven Supply Chain with Axia Value rated in the optimizer model.
Chain Consulting and SAP’s APO,” AMR
Research, November. Document available at
The company SKUs were modeled
www.amrresearch.com. directly (not aggregated) and numbered
6. Tohamy, N. and K. Carter, 2009, “S&OP
Technology Market Update: Picture Still Murkey,” 600 to 1,000. As a result, some versions Introducing INFORMS Conti
Continuing Professional Education Database
AMR Research, November. Document available at
of the model involved more than one Your one
one-stop
stop shop for O
O.R.
R semina
seminars, short courses, and workshops
www.amrresearch.com.
offered throughout the world.
7. Wallace, T. and R. Stahl, 2008, “Sales and Operations million constraints and several million
Planning: The How-to Handbook,” Have a course you would like to add? We want to hear from you!
3rd edition, T. F. Wallace & Co. variables and required several hours to
INFORMS is committed to supporting the lifelong learning of operations research and analytics professionals.
be optimized. Modeling products at the
A^Zema<Zk^K^_hkf
AZlIZll^]'GhpPaZm8
Ik^]b\m^]hnm\hf^l3@nZk]^]himbfblf_hk`^g^kZein[eb\Zg]iZmb^gml4^q\^ee^gmikh`ghlbl_hka^Zema
\Zk^lrlm^flZgZerlml'
;R=HN@E:L:'L:FN>ELHG
CONGRESS’ LANDMARK health country who fully comprehends it. This is Executive branch agencies will now Based on the Kaiser Family Foundation
mandated coverage phases in over Over time, premiums may rise more case, but there will be some rationing Standardization of claims rules may
several years, and one major enforce- slowly than they would have without of care. Actually, this happens now, ease the long-delayed transition to
ment mechanism is that people will be the new law. This is because, along with and it will continue to happen the same standard electronic patient records.
compelled to enroll when they seek some high-risk people, a large group of way: some expensive procedures sim- This improvement in information support
care. There are some exceptions to the young, relatively healthy people will also be ply won’t be covered unless and until for medical decisions could, in turn, sub-
mandate, and several states have filed pushed into the insured pool, lowering the their life-saving necessity can be clearly stantially reduce preventable deaths from
lawsuits contesting the requirement. average risk and cost. Economics predicts demonstrated. What will change is that system errors, principally bad hand-overs
that insurance companies will compete for some of these exclusions will become of information among providers. In 2000,
People excluded from coverage be- these desirable customers by offering more more visible and, therefore, potential- the Institute of Medicine, part of the Na-
cause of pre-existing conditions will attractive rates. The creation of health in- ly more widely debated and reviewed. tional Academy of Sciences, published a
now have more and better choices. This surance exchanges, basically purchasing People with more money will continue study estimating that medical errors ac-
provision takes effect immediately and is pools that negotiate better deals for benefi- to have more choices – such as buying counted for nearly 100,000 preventable
one of the clearest consequences of the ciaries, will push the market toward lower the care they want, perhaps even trav- deaths per year in the United States,
act. Insurance companies will be inten- rates – at least that’s what economic theory eling outside the United States to do so mostly attributable to system problems
sively analyzing how this may affect them predicts. Some states, in particular Califor- – than other people. despite reasonable care by all the indi-
and what they can do about it. nia as a result of operations research ana- viduals providing care and related func-
lyst Enthoven’s efforts there, have reported With or without the public option, tions (such as labs and radiology). They
People with good health insurance success with this approach. Still, what will the nation will move toward a single set followed up with a second publication in
will see their premiums rise. Health care happen on a national scale remains to be of rules and procedures for payment. 2001, recommending system improve-
costs have been rising and will continue to seen. The current hodge-podge of different re- ments to address these problems.
do so; premiums reflect this. Some of the quirements for claims processing is one of Another aspect of uncoordinated care
people who will now get coverage because Some taxes will rise, including some the big reasons that the American Medical is polypharmacy, the use of multiple pre-
of the new law are high-risk and will be put not yet specified. Again, this would also Association and the American Hospital As- scription medications in combination,
into a temporary high-risk pool, similar to have happened without this legislation, and sociation, among other providers’ groups, with too little attention to possible in-
the assigned risk pool for automobile insur- it will be difficult if not impossible to deter- supported this legislation. Some hospitals teractions. Polypharmacy is what killed
ance, cross-subsidized by other people, mine whether the new legislation decreas- have to pay claims specialists to file their Anna Nicole Smith and Michael Jackson,
and insurance companies will want to in- es or increases the effect. As the economy claims with as many as 40 insurance carri- both of whom apparently took deliberate
crease their revenues accordingly. improves, deficit reduction will become in- ers, all with different rules and procedures. steps (multiple doctors, multiple pharma-
creasingly important and politically attrac- Standardizing will save serious administra- cies) to sidestep warnings. Polypharma-
Ln[l\kb[^mh:gZermb\l tive, and Congress will respond. tive costs. As with Medicare now, standard- cy also most likely killed Brittany Murphy,
izing rules and procedures can occur while who does not appear to have received
It’s fast, it’s easy and it’s FREE!
There won’t be “death panels” mak- private insurance carriers continue to ad- adequate warnings or to have ignored or
Just visit: http://analytics.informs.org/
ing life-or-death decisions case by minister the claims. evaded them. Safety testing is usually
done one medication at a time, so inter- treatment quickly from providers who United States alone. No conceivable at- One thing widespread early screen-
actions can take quite some time to be- are part of the reporting network. A de- tack in the United States with a single ing will not do is reduce violent crime,
come identified and publicized. This is a lay of as little as a week in detecting an thermonuclear weapon could kill nearly as the false positive rate of screening is
growing problem, and information tech- incipient epidemic can make a huge dif- as many people. Earlier detection and higher than the incidence of the relevant
nology offers a promising answer. ference in the extent of its spread. The response is the best defense against disorders. On the other hand, more ef-
magnitude of the effect of having most such a catastrophe. fective use of information to keep peo-
There will be much more research people seek medical attention quickly in ple with a known propensity to violence
and assessment about comparative ef- the event of infection is difficult to mea- There will probably be secondary from acquiring weapons would make a
fectiveness of treatments. This is the sure – another analytical challenge. By and indirect economic benefits. Un- big difference. The 2008 Virginia Tech
big payoff for analysts who have taken way of illustration, however, the Centers der the current situation, many people are shootings, for example, could have
the time to learn about medical out- for Disease Control (CDC) estimates forced to base job and business decisions been prevented if the intent of the exist-
comes evaluation, assessment of ef- that, through Feb. 13, 2010, the 2009- heavily on availability and quality of health ing state law had been carried out: the
fectiveness and risk of new treatments, 2010, H1N1 “swine” flu had infected 42 insurance for themselves and their family shooter was omitted from the “no gun
preventive care and other such topics. million to 86 million Americans, mostly members. Removing this constraint should purchases permitted” database after a
The act specifies that the federal gov- before an effective vaccine could be stimulate productivity. Again, measuring the court finding that he was disposed to
ernment will fund more research in this manufactured and distributed in suffi- actual effect will be a challenge. violence because of an ambiguity in the
area. Providers, regulators and insurers cient quantity. This flu, fortunately, had law specifying who had to be entered
will want to know what really works. Liti- unusually low lethality, killing less than Improved access to mental health into the list.
gation over “rationing” will hinge on this .05 percent (8,520 to 17,620) of those care and in the effectiveness of such Many of the commission’s recommen-
question. Employers will want to know infected. More serious pandemic influ- care will produce substantial benefits. dations were formally adopted but not
how to tailor the health insurance pack- enzas within the last century have had The presidential commission that stud- funded to the recommended levels. The
ages they offer to get good perceived lethality of 15 to 20 percent or more; ied mental health care during the last new legislation will direct more money to
value for restrained cost. according to the World Health Orga- administration was emphatic in conclud- mental health, and removing the risk of
nization (WHO), the H5N1 “bird flu” a ing that parity between mental health losing insurance coverage may prompt
National security will improve couple of years ago killed at least 292 of claims and those for other ailments some more people to seek treatment.
somewhat. The best defense against a the small number (493) of people con- would improve health and lower costs. Again, however, measuring actual effects
pandemic is prompt detection, which re- firmed infected, mostly people handling Among the probable benefits are low- and assessing policy alternatives will re-
quires the first people affected to seek birds in Asia – that is, 60 percent lethal- er costs in the criminal justice system: quire quite a bit of analysis.
ity. From these data, it is evident that a most serious mental health diagnoses
A^eiIkhfhm^:gZermb\l flu as infectious as the recent H1N1 but are now made there, as the police and The Democrats will suffer a back-
with 20 percent lethality is a real possi- courts have to deal with large numbers lash in the 2010 elections, but the lon-
It’s fast and it’s easy! Visit:
bility, albeit improbable, and such a flu of non-violent nuisance offenders who ger-term political consequences are
http://analytics.informs.org/button.html
could kill 8 to 17 million people in the should be treated, not incarcerated. yet to be determined. Many people in
the country are palpably uneasy about recession, and there have been no ma- REFERENCES
1. John M. Barry, “The Great Influenza: The Story of the Deadliest Pandemic in History,” Viking Penguin, 2004;
this legislation. Recent polls have dis- jor scandals, social upheavals, or major Penguin paperback, 2005.
approval leading approval of the new military or foreign policy setbacks, the 2. Donald Berwick, E. Blanton Godfrey and Jane Roessner, “Curing Health Care,” Jossey-Bass, 1990.
health plan by margins ranging as high Thirteen Keys model predicts President 3. Centers for Disease Control, U. S. Department of Health and Human Services, www.cdc.gov/flu/h1n1flu/
estimates_2009_h1n1.htm , retrieved April 12, 2010.
as 20 percent while overall disapproval Obama’s re-election, and one can eas-
4. Committee on Quality of Health Care, Institute of Medicine, “To Err Is Human: Building a Safer Health Care System,”
of Congress is as high as 75 percent. ily imagine him asking on the campaign Washington, D.C.: National Academy Press, 2000.
However, even if the Democrats lose trail, “So what was all that fuss about?” 5. Committee on Quality of Health Care, Institute of Medicine, “Crossing the Quality Chasm: A New Health System for
the 21st Century,” Washington, D.C.: National Academy Press, 2001.
control of at least one house of Con- and gaining accordingly. By then, some
6. Tom Daschle, with Scott Greenberger and Jeanne Lambrew, “Critical: What We Can Do About the Health-Care Crisis,”
gress in 2010, the news for President astute congressmen and senators will St. Martin’s Press, 2008.
Obama’s re-election campaign in 2012 have gained credit for good constituent 7. Alain C. Enthoven and Laura A. Tollen, “Competition in Health Care: It Takes Systems to Pursue Quality and
Efficiency,” Health Affairs, Sept. 7, 2005
is not necessarily bad. The same polls service by helping people deal with the
8. Alain C. Enthoven, “Choice in Health Care: Commentary,” Health Affairs, Vol. 25 no. 2, page(s) 566-67,
cited above show his approval and dis- new system and will therefore campaign March/April 2006.
approval ratings about equal. in support of the new law. 9. Laurie Garrett, “The Coming Plague: Newly Emerging Diseases in a World Out of Balance,” Farrar, Straus, and
Giroux, 1994; Penguin paperback, 1995
President Clinton was re-elected com-
10. Kaiser Family Foundation, “Summary of New Health Reform Law,” http://healthreform.kff.org , retrieved
fortably in 1996 despite failing to enact There will be much additional legis- March 26, 2010.
his health plan and losing both houses lation “adjusting” and “correcting” fea- 11. Allan J. Lichtman, “The Keys to the White House: A Surefire Guide to Predicting the Next President,” 2008
Edition, Rowman and Littlefield Publishers, 2008.
of Congress in 1994. According to one tures of this act, as its effects become
12. New Freedom Commission on Mental Health, “Achieving the Promise: Transforming Mental Health Care in
especially successful predictive model – clearer. This is what Congress does to all America,” Final Report, DHHS Pub. No. SMA-03-3832, Rockville, Md., 2003. Also available online from
www.mentalhealth.samhsa.gov.
political historian Allan Lichtman’s “Thir- major legislation. This is most likely the
13. RealClearPolitics, various polling reports, especially http://www.realclearpolitics.com/epolls/other/obama_
teen Keys” – having made a major policy safest prediction we can make. and_democrats_health_care_plan-1130.html , retrieved April 13, 2010.
change helps the incumbent, regardless 14. Douglas A. Samuelson, “A Dose of O.R.: Analysts Explore Ways to Improve Quality and Productivity in Health
Care,” OR/MS Today, December 1991.
of how popular the change is; just having Health care systems analysts will
15. Douglas A. Samuelson, “Diagnosing the Real Health Care Villain,” OR/MS Today, February 1995.
had the clout and leadership skill to get prosper. We all agree that the new law is
16. Douglas A. Samuelson, “A New Frontier? Health Services Research and Medical Informatics,” OR/MS Today,
it through is a favorable indicator. (Pro- complicated. Therefore, there’s a market for February 2000.
fessor Lichtman’s model is admittedly figuring out what it will do and how to deal 17. Douglas A. Samuelson, “Can Early Screening for Mental Disorders Reduce Criminal Justice Costs?” George
Mason University Civil Rights Law Journal, 2001.
controversial. It does have, however, the with it – enough of a market to keep many of
18. Douglas A. Samuelson, “Can O.R. Help Stop ‘The Invisible Plague’?” OR/MS Today, June 2004.
significant selling point of having predict- us employed for a long time. ❙
19. Douglas A. Samuelson, “Can We Detect ‘The Coming Plague’?: How Emerging Health Threats Are Sneaking
ed correctly, well in advance, every presi- Douglas A. Samuelson (samuelsondoug@yahoo.com) is Up on Us,” OR/MS Today, June 2008.
dential election since 1980.) a senior consulting analyst for IBM, focusing on national 20. Jon Stewart, “Commentary: Alain Enthoven: An Outspoken Champion for the Prepaid Group Practice,” The
security and defense issues, and president of InfoLogix, a Permanente Journal, Vol. 8, No. 3, Summer 2004.
If, by 2012, the health care legislation small consulting and R&D company in Annandale, Va. He 21. E. Fuller Torrey, “The Invisible Plague: The Rise of Mental Illness from 1750 to the Present,” Free Press, 2002.
doesn’t look so bad, there is no serious is a frequent contributor to OR/MS Today and Analytics.
22. World Health Organization, www.who.int/csr/disease/avian_influenza/country/cases_table_2010_04_09/en/index.
He has a doctorate in operations research from George
challenge from within his own party or Washington University.
html , retrieved April 13, 2010.
AhpFZma^fZmb\ZeFh]^el<Zg
A^eiBfikho^<Zg\^kMk^Zmf^gmIeZgl
;k^Zdmakhn`aikhc^\m\hf[bg^l^qi^kmbl^bgmphl^iZkZm^Zk^ZlÇkZ]bZmbhgiarlb\lZg]himbfbsZmbhg'
;R@BGHC'EBF
CANCER IS THE second leading A NCI report recently announced we cannot be content with this steady common cancer treatments are sur-
because the organs are often located dimensional conformal radiation ther- from all over the world to block cancer The project proved to be an excel-
in very close proximity to the tumor. apy methods belong to this category, cell growth by inhibiting specific mole- lent example of how expertise in two
Designing a cancer treatment plan of- and they include Gamma Knife radio- cule pathways needed for cells to pro- separate areas – radiation physics and
ten involves a series of optimization surgery, intensity modulated radiation liferate. Radiation therapy techniques optimization – can lead to results that
problems with different kinds of objec- therapy (IMRT), traditional three-di- have become very precise in achieving could not be achieved with expertise
tives and constraints. Researchers in mensional conformal radiation therapy the treatment goals. Engineers work in one area alone. From the optimiza-
the optimization community have made (3DCRT), tomotherapy, intensity mod- tirelessly to develop cutting-edge de- tion perspective, the primary goal of
significant contributions in improving ulated arc therapy (IMAT) and inten- vices to aid cancer treatments. Power- a treatment plan is to deliver the pre-
the quality of various treatment plans sity modulated proton therapy (IMPT). ful and sophisticated cancer scanning scribed dose of radiation to the tumor
for cancer patients such as surgery, The other radiation delivery method is devices have helped improve the detec- while minimizing the dose to non-tumor
radiation therapy and chemotherapy. called brachytherapy. Brachy means tion of cancer. Robotic arms are being cells. The total amount of radiation for
As the oldest form of cancer treat- “short distance” in Greek. Therefore, used in surgeries. Government entities a complete radiation treatment varies
ment, surgery is often used to remove brachytherapy is a term of internal ra- and private foundations have increased depending on the tumor type and the
localized tumors. In cases where the diation therapy where radioactive sub- awareness of cancer, while public health radiation delivery device being used.
cancer has not metastasized, surgery stances are placed inside or close to initiatives have helped induce life style For a typical prostate cancer treat-
may be performed with curative intent. the tumor. changes that reduce the risk of cancer. ment using IMRT, two gray (Gy) of ra-
Radiation therapy uses different types Brachytherapy is often used for diation are administered each day and
of radiation particles such as photons, treating patients with prostate, cervi- 6)7)%6',-26%(-%8-32 the treatment continues up to 40 days.
electrons and protons to sterilize the cal or breast cancer. Effective plans of 8,)6%4=40%22-2+ The reason for this fractionated radia-
tumor or prevent cancerous cells from radiation treatments require delivery of I HAVE BEEN fortunate to work in this tion delivery is based on cell growth
proliferating. Two types of radiation high dose of radiation to the tumor while exciting and rewarding area of research. in the body. Cancer cells reproduce
delivery methods are often used for minimizing the impact on normal cells As a graduate student in optimization at much faster than normal cells. Radia-
radiation therapy. The most common as much as possible. the University of Wisconsin-Madison, tion therapy specifically targets them
radiation delivery machines belong to Drug therapies such as chemotherapy my thesis advisor Michael Ferris intro- by impairing or delaying this cell re-
the external radiation therapy or tele- use medicines to control cancer systemi- duced me to a radiation therapy plan- production process. Normal cells can
therapy. In teletherapy, the radiation cally. Such systemic therapy may be ad- ning problem for brain tumor patients. quickly recover from radiation damage
source is positioned at some distance ministered alone or in combination with The problem was originally presented overnight when they are exposed to a
from the patient. Popular three- other treatments. to him by David Shepard, a graduate low radiation dose. But, if they are ex-
Finding a cure for cancer has been student in Professor Rockwell Mackie’s posed to a high radiation dose, it may
Ln[l\kb[^mh:gZermb\l an exciting goal in research commu- group in the Radiation Physics depart- cause long-term damage. Cancer cells,
nities. Researchers now have a much ment. Shepard’s initial modeling ef- however, are very sensitive to radiation
It’s fast, it’s easy and it’s FREE!
better understanding of cancer. Many forts gave us the groundwork for our exposure and two Gy per day can ef-
Just visit: http://analytics.informs.org/
innovative drugs are being developed research. fectively stop them from replicating.
portion belongs to the located in a patient’s head. It can de- different sizes in a traditional treatment
gross tumor volume (GTV). liver a single beam of high dose radia- planning. Therefore, treatment param-
But it is often difficult to tion by 201 Cobalt-60 unit sources. All eters include a set of radiation shot
draw a clear line between 201 beams simultaneously intersect at center locations, a discrete set of col-
the tumor and the normal the same location to form an approxi- limator sizes and radiation exposure
cells. Therefore, a clinical mately spherical region that is typical- time for each radiation shot.
target volume (CTV) is de- ly termed a shot of radiation. Multiple We formulated this problem as
fined, and it includes GTV shots are often used in a treatment us- a mixed integer nonlinear program-
and its surrounding suspi- ing a Gamma Knife due to the irreg- ming model and solved it as a variant
Figure 1. (Left) Multiple angles are used to achieve a high radiation dose on cious area. Finally, plan- ularity and size of tumor shapes and of the sphere-packing problem where
the tumor while minimizing radiation exposure to surrounding normal cells. ning target volume (PTV) the fact that the focusing helmets (or the PTV can be viewed as an empty
is constructed for treatment collimators) are only available in four container. Next, our job was to find
Figure 2. (Right) Beam’s-eye-view is constructed for each angle.
planning. PTV adds mar-
Radiation treatment planning begins gins to CTV to account for
with a set of images of the organs of movement of organs during treatment
interest. Such images can be obtained and the movement of the patient be-
by diagnostic imaging tools such as tween treatments.
computerized tomography (CT), mag- The images are further divided into
netic resonance imaging (MRI) or posi- a collection of three-dimensional cubes
tron emission tomography (PET) scans. (or voxels) for optimizing treatment pa-
Your one-stop shop to view top presentations from key INFORMS meetings
Those diagnostic images then help phy- rameters. This is where optimization
sicians determine the three-dimension- models can play a key role in treatment
al shape, location planning. The type of radiation delivery 2009 Annual Meeting
and size of the tu- device determines what parameters Plenary
mor and surround- need to be optimized. My colleagues Keynote
Wagner Prize Presentations
ing normal tissues. and I have developed optimization
Delineating treat- models for Gamma Knife radio-surgery, 2010 Practice Conference
ment volume on the IMRT and IMPT. Edelman Award Presentations (coming soon)
images is often a
difficult task to per- +%11% /2-*)6%(-3796+)6=
Your latest member benefit lets you learn from the best on your schedule.
form. A well-trained T HE G AMMA K NIFE IS a stereotac- http://livewebcast.net/INFORMS_Video_Learning_Center
Figure 3. Varian Medical radiation oncologist tic radiation delivery device for treat-
System’s multileaf collimator. can identify which ing tumors or vascular malformations
a mix of spheres to fill the container for a variety of cancers. In 3DCRT, solved using an interior point method. The potential of IMPT is only now be-
while minimizing the total space that multiple external radiation beams are Since the MIP model contains more ing fully understood, and optimization
was not covered. Another goal: mini- directed from different angles into the than one million continuous variables is an important part of realizing IMPTs
mize the number of spheres to do the tumor (see Figure 1). Each beam can and several thousand binary variables full potential. I am presently working
job. We have successfully developed a be shaped by a computerized multi- and more than one million constraints, in collaboration with the University of
computer program, tested it and used leaf collimator mounted on a gantry to solving the problem optimally using Texas MD Anderson Cancer Center in
it in radiation planning for patients at conform the beam’s-eye-view of the B&B can take several weeks using a Houston on optimizing the delivery of
the University of Maryland hospital in tumor (see Figure 2). Hence, a lethal fast workstation PC. Therefore, heu- proton therapy. MD Anderson Cancer
Baltimore. The reliability of this treat- dose of radiation can be precisely de- ristic approaches such as local neigh- Center started treating certain types of
ment-planning tool has been proven by livered to a tumor. Photon beams are borhood search, simulated annealing cancer patients at their Proton Center
clinical trials at the hospital. In eight out used for 3DCRT. and genetic algorithms are often de- in 2006.
of 10 cases, the tool produced plans As an advanced form of 3DCRT, veloped to expedite the treatment plan-
that are better than the ones made by a IMRT utilizes an advanced multileaf ning process. '32'097-327
physician. Both the speed of the treat- collimator that can move its leaves C ANCER IS AN abhorrent disease, but
ment plan generation and the treat- back and forth to block the portion of -28)27-8=13(90%8)( it’s a disease we continue to make
ment quality are the main contributions each beam’s radiation dose (Figure 4638328,)6%4= strides against every day. One way is
of this decision-making tool. 3). This enables a high degree of flex- PROTONS WERE first proposed for radiation through cross-disciplinary efforts, such
ibility in delivering radiation from each therapy by Robert Wilson at U.C.-Berkeley as bringing the powerful tools of op-
-28)27-8=13(90%8)( gantry angle. Typically, between five in 1946. Since then, thanks to more than timization to bear on problems in ra-
6%(-%8-328,)6%4= to nine angles around the 360-degree five decades of continuous research and diation therapy. But the application of
IMRT IS A radiation delivery method circumference are selected for the development, we now have a potentially optimization, and analytics in general,
under the umbrella of three-dimen- treatment. Radiation is delivered at a very powerful radiation delivery machine isn’t limited to cancer treatment. It can
sional conformal radiation therapy. fixed angle, and it continues until all that uses proton beams. It has been report- be applied to the design of new drugs,
3DCRT utilizes 3-D imaging data of the selected angles are covered. This ed that proton therapy may be more pre- imaging and genetic sequencing, not
the tumor and surrounding anatomy to treatment-planning problem is often cise than other forms of radiation treatment to mention hospital planning, schedul-
achieve a high conformal dose of ra- formulated as a mixed integer linear in certain types of cancer. This is because ing and administration. Medical prac-
diation delivered directly to the tumor. programming (MIP) model, in which a the dose distribution from a mono-energet- titioners from all walks can benefit by
This treatment can be administered set of gantry angles and their radia- ic beam of protons increases slowly, rising adopting the many powerful tools ana-
tion intensities are optimized. In our to a sharp peak, called a Bragg Peak, at lytics has to offer. ❙
A^eiIkhfhm^:gZermb\l IMRT treatment-planning tool, the MIP the target. By superimposing beams of dif- Gino J. Lim (ginolim@uh.edu) is an assistant professor
model is solved using the Branch-and- ferent energies, IMPT does only moderate of Industrial Engineering at the University of Houston.
It’s fast and it’s easy! Visit: The author acknowledges Andy Boyd for his valuable
Bound (B&B) technique, and the linear damage to tissues on its way to a tumor comments and suggestions in preparing this article.
http://analytics.informs.org/button.html
programming relaxation models are and almost no damage thereafter.
IKBG<>MHG<HGLNEM:GML printer and scale it up for a company the size of changing their plans at the last minute. Weather Allocating space: Space is another asset
<K>:M>=:GHIMBFBSBG@ Quad/Graphics. Optimizing for Quad/Graphics and air-traffic patterns are another source of con- that can be as scarce as equipment and no
LH?MP:K>IKH@K:FMA:M required building optimization software from the stant uncertainty and change. So choosing when less contentious. Take ad space, for example.
MHHDBGMH:<<HNGM ground up. and where to fly each jet and crew at the moment It is a precious and limited asset, both for busi-
:EEMA>KNE>L:G= of truth – when a customer calls and wants a jet nesses needing to reach a target audience and
8,)6)79087 now – is an energy-draining challenge. for those who own the media outlet, making ad
<HGLMK:BGMLNG=>K
PAB<AMA>:BKEBG> IN ITS 29-YEAR HISTORY, Princeton Consul- Princeton Consultants’s optimization team space a prime candidate for optimization. The
tants has helped clients reap large benefits by successfully created an optimizing software Wall Street Journal provides a case in point.
HI>K:M>L:G=IK>L<KB;>=
applying the principles of optimization to vari- program that took into account all the rules Given the complexity of its numerous edi-
MA>HIMBF:ELHENMBHG'
ous areas. Following are examples of its work and constraints under which the airline oper- tions and the number of rules governing where
with clients in three major ones: scheduling, ates and prescribed the optimal solution. The ads can be placed, laying out the paper’s ads
allocating space and pricing. bottom-line results? NetJets successfully uses is tricky business. For many years, it was done
Princeton Consultants’ models in production to manually each night by two employees. The
Scheduling: For many high-profile person- schedule all worldwide pilots and jets. ads that didn’t fit were shelved, and advertisers
alities who shun commercial flights, the airline
of choice is NetJets. NetJets flies approximate-
ly 400,000 flights annually on 775 aircraft to
approximately 175 countries around the globe.
Each of NetJets’ well-heeled passengers is a
part-time “owner,” for whom a plane is avail-
able whenever and wherever it is needed. The
resulting complexity makes efficient schedul-
ing a major challenge.
Knowing that improvements in fleet utilization
would give its owners tremendous advantages,
NetJets called on Princeton Consultants to create
a scheduling system that would maximize utiliza-
tion while minimizing customer dissatisfaction.
Many rules govern the decision as to who
should get which jet, when. Some are related
to the size and speed of the planes. Others are
mandated by the FAA. There are also union
regulations to consider and company policies.
Another complicating factor: The high net-worth
individuals who use NetJets often think nothing of
:L:K>LNEMH?MA> didn’t learn that they had been left out until the could give them a price quote in seconds in- successes, other industries have been slow to
IKHC><M%MA>P:EELMK>>M next morning. The result: lost revenue and lost stead of days? What would such agility do for follow their lead. For example, the healthcare
CHNKG:E=B=GHMA:O> advertisers, who headed for more reliable ad- your company’s reputation? How many more industry has been slow to adopt optimization
MH;NRFHK>IK>LL>L% vertising venues. prospects would call for a quote? techniques. Much of the discussion concern-
ABK>FHK>I>HIE>HK Realizing that it was time to replace this anti- One of Princeton Consultants’ clients is ing optimization in healthcare represents aca-
=BL<HNGMBMLK:M>LMH quated system, the paper’s management asked CSX Railroad. Each day, CSX moves more demic research, not field applications used on
BG<K>:L>BMLK>O>GN>' Princeton Consultants to develop a program that than 20,000 railroad cars of essential items a daily basis to reduce inefficiencies and im-
would optimize WSJ’s use of advertising space. throughout North America. Some shipments prove service. Paul O’Neill, former secretary of
The firm’s team worked side-by-side with the are scheduled months in advance; other re- the Treasury and veteran optimizer, summed
employees who had created the playbooks and quests come in at the spur of the moment. up the extent of the missed opportunities when
about 100 support staff at the paper to design The employees who schedule the freight ship- he said, “For me, there is no bigger smorgas-
software that rationalized the ad-placement pro- ments need to know, at all times, exactly how bord of potential opportunity for the application
cess and drove up value. much space is available, on which trains and of your [optimizers’] talents and insights than
As a result of the project, the Wall Street how much it can be sold for. If a CSX train from in the practice of American medical care.”
Journal did not have to buy more presses, hire Jacksonville, Fla., to Chicago is going through It is not just the medical profession that has
more people or discount its rates to increase Nashville, Tenn., with several empty cars, and been slow to adopt optimization. Throughout the
its revenue. Just getting more ads into the pa- a business in Nashville can fill a couple of business world, companies continue to manage
per each day did it. Placing ads with the Wall them, CSX can probably offer the shipper a scarce and expensive assets in traditional ways,
Street Journal also became much more cus- favorable price. The optimization software cre- shunning proven optimization techniques.
tomer friendly. Now, an advertiser can call and ated and installed by Princeton Consultants Sashihara maintains that this is about to
ask for an ad and find out almost immediately can calculate the price point that will make it change. “I am very excited about the future of
if it will appear in the next day’s paper, elimi- a win-win for both CSX and the customer, and optimization for two reasons,” he says. “The
nating surprises. The bottom-line results? Dow it can do so instantly. The bottom-line results? plummeting cost of hardware and globaliza-
Jones & Company successfully uses Princeton CSX successfully uses Princeton Consultants tion. You can’t just cut costs, centralize and
Consultants’ models in production to layout all models in production to provide instant Web mass produce anymore. You need to be able
worldwide editions of the Wall Street Journal. pricing and order booking, making what CSX to ‘mass customize,’ which is what optimiza-
Intermodal President James Hertwig called “a tion enables you to do.”
Pricing. Agile price setting can be a com- prime example of our determination to be the Sashihara, who also serves as co-chairman
petitive advantage is many industries, espe- most progressive intermodal service provider of the Association of Management Consulting
cially in large, B-to-B transactions. How would in North America.” Firms (AMCF), adds that optimization is about
your customers and prospects react if you to emerge as one of the hottest growth areas
8,)*9896)3*348-1->%8-32 in consulting. ❙
Ln[l\kb[^mh:gZermb\l SOME INDUSTRIES, such as transportation, Peter M. Tobia, Ph.D. (pmtma@aol.com) is president of Market
It’s fast, it’s easy and it’s FREE! Just visit: http://analytics.informs.org/ have been using optimization for decades with Access, a marketing firm and literary agency based in Wyndmoor,
Pa. He has written for many major business publications.
excellent results. Despite their demonstrated
IZmb^gmGh'+*
Making a correct diagnosis on a patient is
a very challenging task for a doctor. A complex
disease can manifest through any number of
symptoms that often leave “trial and error” as
the only method for arriving at the correct di-
agnosis. To aid doctors in this daunting task,
mathematical modeling can be used to help
guide a diagnosis.
Table 1 shows 20 patients with varying symp-
toms and corresponding diagnoses. For exam-
ple, Patient No. 1 reported symptoms of sneezing
and a sore throat and was found to have a cold.
Patient No. 20 exhibited symptoms of fatigue
and sneezing that was found to be allergies. All
diagnoses were confirmed through lab tests.