Izabela Annis

david_and_izabela_annis.jpg

M.S. Statistics 2001

Senior Statistician, Nielsen Analytic Consulting

Written by: Shannon Knapp, Ph.D. candidate in Statistics 

As a market research statistician Izabela Annis enjoys that her clients make decisions based upon her analyses.  She works for Nielsen, the company popularly known for their television ratings.  But Nielsen is more than TV ratings; in fact, they are the industry leader in market research, quantifying the performance of products and advertising to guide clients' marketing strategies. 

Izabela came to Nielsen because she wanted to do applied work and address real world problems.  "They showed me the models, talked about the data.  It looked like I could make an impact."  When she started with Nielsen, she ran many of their standard analyses. Two such products are marketing mix, which quantifies the impact of marketing elements (TV ads, magazine ads, coupons, store displays, price cuts) to maximize sales, and market segmentation, which clusters households based on purchasing patterns.

Izabela has since worked as part of an elite team designing and implementing custom analyses for high profile clients.  Recently she worked with a prominent carbonated beverage company to investigate a potential link between brands available at restaurants (restaurants typically offer only one brand of carbonated beverages) and those consumed at home.  In another case, a client was interested in "trip-segmentation," clustering not by households, but by shopping trips to determine whether their product was purchased during weekly grocery trips or quick runs to the store to get a few items.  She has used a variety of statistical techniques, including multinomial logit, time-series, clustering, discriminant analysis, and generalized linear models to answer these and other questions for her clients.  "Every new problem is a new challenge."

Izabela has access to data from three major sources: food and drug store sales, advertising, and household panel data.  But what she uses the most is the household panel data.  Thousands of consumers selected to be on this panel scan every item they purchase, report if they used a coupon, and the price they paid. With data on every purchase made by thousands of consumers over many years, she has become comfortable dealing with large data sets.  "Real world data can be very messy.  It's amazing how long it takes to clean up your data."  Izabela strongly recommends statistics students get an internship, if only to appreciate the scale and problems of real datasets.  "You will never see 'textbook' datasets with 50 observations in practice."

When Izabela's husband, David (M.S. Statistics 2001, Ph.D. Statistics 2003), found a position in California in 2004, her boss arranged for her to keep her position by telecommuting, both because they valued her abilities and because they trusted her work ethic.  "I'm very self-disciplined.  I get the work done.  I've always been this way."  She travels to her Chicago-area office every 3 months for a week at a time for team meetings or training.  While she enjoys the lack of a daily commute and the flexibility of her time, she admits she can feel quite isolated.  She misses the convenience and informality of just being able to walk over to a colleague's desk and say, "Hey, what do you think about this?"

Izabela came to the Purdue Statistics Department for its reputation.  "I wanted to study under the best, and it turned out to be true.  The graduate students and professors at Purdue are among the smartest people I have ever met."  Although there is a lot of learning on the job, at Purdue, "I learned how to learn."  She also appreciates her teaching experience at Purdue.  Being a teaching assistant for STAT 113 (Statistics and Society) "was very helpful learning how to explain concepts and relevance to people without the technical background."

Working in market research also gives her a window into the oddities of consumer behavior.  "In this field everyone has a story."  Hers came while studying toilet paper.  One of the typical measurements done with any product is penetration: the percent of households that had purchased the category at least once.  With two years of purchase data, one would assume nearly every household had purchased toilet paper at least once.  But the penetration rate for toilet paper was only 92%.

To read more Alumni Profiles, please visit our Alumni Profiles archive.    

Last Updated: Aug 29, 2017 4:27 PM

Purdue Department of Statistics, 250 N. University St, West Lafayette, IN 47907

Phone: (765) 494-6030, Fax: (765) 494-0558

© 2015 Purdue University | An equal access/equal opportunity university | Copyright Complaints

Trouble with this page? Disability-related accessibility issue? Please contact the College of Science Webmaster.