GBK Employee Spotlight: Dan Yavorsky
GBK Employee Spotlight is a series designed to surface the stories of the amazing individuals across our team and what makes them tick.
Today for our employee spotlight series, we chatted with Dan Yavorsky, PhD, who recently joined as Senior Vice President, Analytics and Marketing Science, in our LA office. Dan brings more than 10 years of analytics and consulting experience including strategic management at Bain & Company, and prior to that, Cornerstone Research.
In his new role, Dan oversees GBK’s analytics & marketing science practice, advising clients in the areas of customer segmentation, choice modeling, customer lifetime value, brand equity, marketing attribution and marketing mix modeling, among other areas. Let’s learn more about him.
Q: Tell us a bit about yourself, your prior career in analytics, and how you got started at GBK.
I started my career doing analytic work at Cornerstone Research, a firm that provides economic and financial analyses to support expert witnesses in corporate litigation. At Cornerstone, I worked under the direction of business school professors and alongside PhD-holding economists. They demonstrated how advanced statistical methods could quantify the causal impact of a firm’s past actions. I was hooked. I knew in an instant that my career would forevermore involve data analytics and econometrics.
To build a toolkit of econometric methods, I pursued a PhD in quantitative marketing at UCLA’s Anderson School of Management. My research combined (1) a structural model of consumer behavior from economics, (2) a unique dataset of millions of cell phone geolocations, (3) estimation methods involving simulation developed by statisticians and computer scientists, and (4) a managerial question about consumer decision-making.
My research studied consumers’ preferences in the automotive market, and how a marketing program – such as the offering of at-home test drives – might influence where consumers shopped and what they chose to buy. I demonstrated the differential effects of such a program across the top automotive brands, and how advanced-modeling techniques that incorporated consumers’ shopping patterns improve on the traditional discrete choice models that are commonly applied to these types of problems.
After graduate school, I joined Bain & Company’s Advanced Analytics Group where I put my toolkit to work, helping companies make informed, strategic marketing and business decisions. Our clients spanned consumer packaged goods, health care products, media and entertainment, among other industry sectors. We used time-honored techniques in segmentation and discrete choice, as well as newer embeddings-based machine learning models of consumer behavior.
I thoroughly enjoyed the work at Bain; however, GBK offered me an amazing opportunity that I simply had to pursue. In particular, the GBK model enables me to collaborate with leading academic experts to employ cutting-edge analytic methods to support industry leading clients. I am currently working with GBK co-founder Professor Eric Bradlow on several projects – which has been fantastic – and I’m looking forward to future projects with our other academic partners. GBK is also supportive of my desire to keep one foot in the classroom: I currently teach Customer Analytics at UCSD and Econometrics and UCLA. I believe the consulting and teaching opportunities are complements, each improving the other, and I look forward to a long career at GBK with this balance.
Q: What are the biggest emerging trends in analytics, or methodologies you are most excited to develop further through your work with client partners?
There are several! I will mention three:
1. Prediction. Datasets are larger, computers are faster, and computer science has contributed new tools for prediction. As a result, we can easily employ flexible model specifications and make predictions that are driven more by the data than by our models. For example, what once was a linear regression might now be a random forest, XGboost, or neural network model. Let’s use the newer, more-flexible tools where appropriate.
2. Causal inference. Biostatistics (with randomized controlled trials) and economics (emphasizing endogeneity) have long used tools for counterfactual reasoning. These tools are now being adopted by marketers and business strategists. Propensity scores, instrumental variables, and regression discontinuity designs are underused. Let’s apply this machinery to, e.g., study causal relationships between marketing variables and firm outcomes in order to inform optimal deployment of firm resources.
3. New approaches. An understanding of consumer demand is fundamental to informing many marketing strategy choices (and is in-part why conjoint is such a powerful tool for marketers). Recent advances have relaxed the hyper-rational economic assumptions about consumers and offered new approaches to measuring consumer demand (see, e.g., Kim, Albuquerque, and Bronnenberg (2010) “Online Demand Under Limited Consumer Search” Marketing Science 29(6) and Chick and Frazier (2011) “Sequential Sampling with Economics of Selection Procedures” Management Science 53(8).) I’m excited to work with client partners to implement these models.
Q: What’s a fun fact about you that many people may not know?
My set of hobbies is quite diverse, ranging from fairly cool to incredibly nerdy. Roughly in order, we have: collecting scotch and bourbon, cycling, urban hiking while traveling (ie, self-created city tours), collecting traditional pocket knives, amassing a library of statistics textbooks, building custom mechanical keyboards, and proselytizing use of the oxford comma. However, I spend most of my time collaborating with clients, teaching students how to analyze data with R, and raising two kids with my incredible wife.