When the Barnes Foundation had questions about its attendance data, a team of Penn students answered through the Wharton Analytics Accelerator Challenge.

One of the world’s finest collections of impressionist, post-impressionist, and early modernist paintings, the Barnes Foundation experienced a transformative move in 2012 from its longtime home in Merion to a new building on Philadelphia’s Benjamin Franklin Parkway in 2012. Remarkably, the number of members went from 400 to 20,000 by the time they opened in Philadelphia. Attendance more than tripled the first year.

“Our mission as we moved to Philadelphia was to welcome people of all ages, ethnicities, and economic backgrounds,” said Peg Zminda, executive vice president, chief financial officer, and chief operating officer of the Barnes. “The scope of our operation is very different now compared to when we were in Merion.”

But with that broad success came some new challenges.

“When we opened we were the hottest ticket in town,” said Zminda. “What we’ve seen over the past six years is that demand has declined. You expect that from a new cultural institution, but we have not yet figured out what a steady state looks like.”

The Barnes heard about Wharton’s Analytics Accelerator Challenge, submitted a proposal, and was chosen for the data challenge.

Bringing the Classroom to the Real World

Now in its second year, the Analytics Accelerator Challenge, a program of the Wharton Customer Analytics Initiative (WCAI), once again paired student teams with corporate and nonprofit partners to solve real problems with analytical solutions. 200 students and 30 companies applied. The Wharton Data Analytics Club helped choose 24 students and assign them to teams to address the problems of four selected organizations, which, in addition to the Barnes Foundation, were Fuel Cycle, Hachette Book Group, and Reed Smith.

The students were given four weeks to work with datasets provided by the organizations. Their hard work culminated with the final Summit on Friday, November 2, where they presented their findings to an audience of corporate partners, students, and other analytics enthusiasts.

The Barnes student team — Catherine Bache (W’21), Karen Chi (C’19, W’19), Anuj Gupta (WG’20), Huey Han (WG’20), Sid Jaiswal (ENG’21), and Kai Lu (C’21) — included MBAs and undergrads from several Penn schools.

As Mary Purk, executive director of the WCAI, said of the accelerator: “One of the objectives is to bring all the disciplines together to solve problems.”

Quality over Quantity

The goal for the Analytics Accelerator Challenge team was to create a model for predictive analysis, incorporating pricing, revenue projection, visitation and attendance behavior, products, and offers. It was a big task to complete in just four weeks.

The Barnes Foundation had robust development and admissions data systems, but it was difficult to forecast attendance when only the past five out of 95 years of history were relevant for the analysis. Ultimately, the team found they could use the Barnes’ aggregate data for their model.

“We used the ticket data aggregated by month rather than granular raw data because it would be more useful for forecasting,” said Kai Lu, an undergraduate member of the team.

Kai Lu, C’21 (left) with teammate Karen Chi, C’19, W’19

“Since they already process data into aggregated sets, we wanted to give them something they could use easily,” MBA Anuj Gupta explained. “We didn’t want to give them extra work.”

The team forecasted attendance in a “do-nothing” scenario and then anticipated various scenarios that could make an impact. They structured two separate models — one for non-members, and another for members and free tickets. They hypothesized that demand and attendance could be influenced by Barnes actions (prices and discounts, special exhibitions) and macro factors (seasonality, market health).

The team determined that in a “do-nothing” model, Barnes attendance would have a 10% year-on-year decline. However, the team found that they could create modest discounts in price that would stabilize attendance and remain revenue-neutral — a 1% reduction in price would have a 1.5% increase in attendance. They also recommended two special exhibitions per year in Q1 and Q3, since seasonal attendance otherwise spiked in Q2 (summer) and Q4 (winter holidays).

Special exhibitions at the Barnes are limited by space.

“We only have one special exhibition space so we can only do one at a time – this is a new activity for the Barnes so going from zero to one felt like enough,” said Zminda, who said they are also looking to create additional venues within the building for special exhibitions.

Still, the student team found that one exhibition at a time was enough.

“It wasn’t so much the number of exhibitions that made a difference,” said team member Catherine Bache. “For non-members, it was whether or not there was one in the quarter. For members, the factor was whether there were members-only special events.”

“The quality of the exhibitions matter more than the number,” said Anuj. “The Picasso exhibition was more impactful than others because of the profile of the artist.”

Anuj Gupta, WG’20, presenting at the Summit

What the Barnes Got out of It

“We’ve been looking for the pricing sweet spot,” Zminda said. On the one hand, a higher ticket price generates memberships because [of] the higher the ticket price, the more attractive the membership. On the other: “Our ticket price has been relatively stable, but we’ve been concerned that since we changed our ticket price in 2016, it pushed us to the top of the price list among our peers. We’ve asked ourselves whether we pushed it too far.”

What the student team discovered surprised the Barnes.

Zminda reflected: “The student model indicates that we could reduce the price and the impact would be revenue neutral. We’ve been thinking hard about the price of our ticket and whether it’s too high, and a model that helps us understand what happens when you lower price would be very useful. In the field, the thinking is that museums are price insensitive but this model suggests the opposite.”

What the Students Got out of It

“It was a cool experience,” said Catherine, who was especially interested in the mission of the Barnes Foundation. “We all had different backgrounds. The MBAs obviously had work experience coming into it. I hadn’t even taken any coding classes, but I was able to learn more and work under their guidance. We were split off into smaller teams to tackle different problems together.” Catherine served as project manager, bringing logistics and organization to the team.

Catharine Bache, W’21

Her teammate Kai, an economics major in the College, brought analytic firepower. “I’m interested into the mathematics of economics,” he said. “I looked into this opportunity to have more of a business focus in analytics. Work on my projects is very individual-based and I don’t have a lot of collaboration, and that’s an important skill set. Catherine’s a great project manager. She made sure we all met on time and communicated and motivated us to put in that extra meter of work.”

Both Catherine and Kai are taking what they learned to their next endeavors.

“The project has shown me how interesting analytics can be and that you can apply it to any industry,” said Catherine. “I’m interested in getting an internship this summer that’s related to data analytics.”

“I’m using what I learned in forecasting demand and how I can apply that to different industries,” Kai said. “Right now I’m working with a friend on how to reduce food waste in the airline industry.”

What’s Next

The Barnes isn’t finished with the data yet.

“I want to back-test data and evaluate the budget for 2019 and see how it fits the budget the students have developed,” said Zminda. “We’ll use it as a check. Going forward, I want to evaluate to see if it’s a tool we use all the time to forecast admissions revenue. It was great to have someone who was able to look at the challenge with a clear eye.”

The Barnes Foundation, which has a historic relationship with Lincoln University, is committed to working with educational institutions in the region.

“Being forward-thinking is important to us, and part of that goes back to our founder who was a forward-thinker in the field of education,” Zminda added. “We see part of our mandate to be a progressive institution. For me, the place where progressive stuff happens is in colleges and universities. They have smart students and resources and think about things in a different way.”

Partnerships like this benefit both the Wharton Customer Analytics Initiative and the Barnes Foundation.

“The most interesting data is sitting out in practice and that’s the foundation of everything we do,” said Prof. Eric Bradlow, faculty director of WCAI, in his introduction to the Summit.

“We’re a relatively small organization relative to our peers. We can’t do this all on our own,” said Zminda. “If we can bring our assets to the table and they can bring theirs to the table, the sum will be more than one plus one.”

Read more on this year’s Analytics Accelerator Challenge Summit and the student team that partnered with market research firm Fuel Cycle.

— Kelly Andrews

Posted: December 6, 2018

Related Content

Read More Stories