As an Amazon Associate I earn from qualifying purchases.

Ozge Sahin on the art and science of studying consumer behavior

[ad_1]

Buy one, get one free: It’s a time-honored shopping offer that’s often a no-brainer for customers. In the operational research world, BOGO is an example of nonlinear pricing. If done right, such pricing can be a win for both consumers and retailers.

Amazon Scholar Ozge Sahin is a professor of operations management and business analytics at Johns Hopkins and an expert in understanding how pricing affects customer decisions.

Amazon Scholar Ozge Sahin explores nonlinear pricing, consumer behavior, and other aspects of business analytics in her work with the Amazon Pricing Research and Machine Learning group. A professor of operations management and business analytics at Johns Hopkins University’s Carey Business School, she is an expert in understanding how pricing affects customer decisions.

Since becoming an Amazon Scholar in August 2019, Sahin has conducted research in two primary areas: bundle promotions and quantity discounts. Each project uses analytical models to determine which discounts or promotions can benefit both customers and Amazon’s retail business.

Bundling and quantity discounts

The bundling project involves nonlinear price discrimination based on an assortment of products. For example, the Amazon Store might display a product grouping and offer a discount if the customer makes multiple purchases within that group. A customer might get 20% off if they buy two or three items within that grouping with the flexibility to build a shopping basket from other items that fall within the promotion.

Product comparison screen.png

Related content

Crowdsourcing reveals sources of information that could improve product discovery algorithms.

Sahin developed an algorithm with an embedded economic consumer behavior model that helps determine the best bundled promotion assortment to offer to customers. If a product sells out, or there’s a new product that would enhance the quality of the assortment and experience of the consumers, the algorithm updates the choices accordingly.

Her second project focused on quantity discounts, or the opportunity for a customer to pay less when they buy more of a specific product. “There are many benefits to quantity discounts,” Sahin says. “Definitely it helps customers save on a per-unit basis, but it also promotes sustainable practices by decreasing the number of shipments.”

Buying six bars of soap at once, instead of one bar a month, for example, not only saves money for the customer but also prevents five additional shipments — and the packaging that comes with them.

An aerial view of the inside an Amazon facility, there are rows of boxes and tables with workers moving among them

Related content

How Amazon’s Delivery Experience team acts as a concierge for customers.

In separate research using data unrelated to Amazon, Sahin has found that bundling is effective at boosting sales, and it has an interesting side benefit: People are less likely to return their purchases, which has the environmental upside of mitigating waste.

Insights on inventory flexibility and buyer decisions

In addition to consumer behavior and pricing, Sahin’s research covers the value of flexibility in supply and demand.

She provides the example of a hotel offering standard rooms, deluxe rooms, and a presidential suite. The different room tiers provide supply flexibility (i.e, hotels can fulfill standard room demand with a deluxe room); while the ability to discount those rooms on an ad hoc basis offers demand flexibility. The two together, she has found, can be an important tool for both businesses and customers, smoothing out price swings.

This picture is an overhead shot inside an Amazon center, workers can be seen moving amidst hundreds of boxes which sit on conveyor belts and carts, in the upper left foreground, a yellow railing extends into the distance.

Related content

How Amazon’s scientists developed a first-of-its-kind multi-echelon system for inventory buying and placement.

“If you have flexibility in your inventory or resources, and you can change prices over time, and you do it in a coordinated, smart way, the price difference between your products will not fluctuate too much,” Sahin says. “So you can offer a more consistent, easy-to-understand product portfolio and pricing menu to your consumers.”

In other research, Sahin explored the question of whether more complex pricing options or structures present challenges for shoppers. Sahin has found that this is not the case.

“Interestingly, what we find is customers are really good at making cost-optimal decisions,” Sahin says of one study that looked at subscription packages versus pay-per-use.

In follow-up research on sequential purchasing decisions made over time, she and colleagues found again that consumers are savvy about picking the best choice — provided they have frequent feedback along the way. She uses an example of a college student on a fixed-cost meal plan: Learning the menu over time and being able to evaluate the choices against one’s own tastes leads to the smartest use of those limited meals.

Balancing academia and industry

At Johns Hopkins, Sahin teaches business analytics and operations management.

“We teach our students how to use analytical models to make smart decisions,” she explains. She is also faculty director of the business school’s Innovation Field Project Course, part of an experiential learning series and a core offering for full-time MBA students. The school works with industry partners to come up with business challenges that students work on for eight weeks with the goal of producing data-driven solutions.

Sahin became interested in operations research while she was an undergraduate at Bilkent University in Ankara, Turkey, where she earned a bachelor’s degree in industrial engineering.

Math is beautiful, but it should be relevant. There is an art to writing analytical models that are tractable and also good representations of reality — we’d like to solve and learn from these models.

The industrial focus introduced the idea of using mathematical tools to make business and economic decisions, and Sahin went on to earn master’s and PhD degrees in operations research from Columbia University. Over the past 15 years or so, she has seen operations research evolve significantly.

An overhead image shows a fulfillment center with boxes and conveyor belts in a large, open facility

Related content

The SCOT science team used lessons from the past — and improved existing tools — to contend with “a peak that lasted two years”.

“When I first started, working with data was rare. The focus was on developing new methodologies and deriving theoretical insights,” she says. “Now the field has moved to a more data-driven approach. We can see the impact of what we are doing in practice and test our algorithms. It’s a change in the right direction.”

Sahin learned about the Amazon Scholars program through a conversation in 2018 with Robert Phillips. At the time, Phillips was at Uber and would soon join Amazon, where he was director of pricing research until 2021. Planning a sabbatical, Sahin mentioned that she wanted to spend the time in industry, learning about the new challenges and consumers’ problems.

“Hearing about the Amazon Scholars program, I thought this would be a great way of knowing how business is done at such a big retailer,” Sahin says. So she joined the company in that capacity in August 2019.

As she looks ahead, Sahin says she likes thinking about the puzzle of how technology can both delight a customer and provide business solutions. But she doesn’t want the process to be just a thought exercise.

“Math is beautiful, but it should be relevant,” she says. “There is an art to writing analytical models that are tractable and also good representations of reality — we’d like to solve and learn from these models.”



[ad_2]

Source link

We will be happy to hear your thoughts

Leave a reply

Discover Your Essential Style at NovaEssentials
Logo