The creative function is what most people think of when you say, “ad agency,” and what attracts people to the field. Although SEO and engaging people through social media are the core competencies, data – if used well – makes a big competitive difference. In an age of all-digital advertising, Jeremy Bice (MBA, ’17) says “everything is data-driven.” Agencies need to crunch the data with great precision to see what ads are working with indicators like cost-per-click, as well as conversions and return-on-ad-spend (ROAS).
Bice started off at Logical Position, a digital marketing agency, as an account manager but had an interest in data that stretched back to his work in signals intelligence in the Marines. He moved from accounts to the paid search team, which relies on Google’s AdWords for its clients’ digital marketing data.
Customer retention is critical to any business, and the agency wanted ways to improve the numbers. “We hypothesized that we might find data hidden in our in-house data and Google AdWords about which clients were in jeopardy and what to do about it,” says Bice. For his efforts and those on his teams, Google gave the agency an award for integrating and leveraging its data.
Discoveries about the Client Life Cycle
Using business analytics skills he learned at Golden Gate University, such as Python and R programming, Bice discovered that there was a relationship between clients’ spend level and the client life-cycle — which spans onboarding to multiyear relationships. The results revealed that clients were most likely to become more engaged or drop at certain stages. The team also discovered that if clients made it past a certain lifespan milestone, their customer lifespan would rise significantly. That’s where services could be bolstered.
Bice and the team also integrated customer interactions with staff with the AdWords data—and did a text analysis on the staff notes with Python. Staff also rated how the calls went on a scale from 1-5, adding another dimension to discover customer engagement.
“…when I look at a data visualization with someone at work,
they immediately say one thing, but I can offer an interpretation based on the business processes as a whole. That is very useful.”
—Jeremy Bice (MBA, ’17)
After standardizing and consolidating data, adding new fields to supplement Google data, Logical Position had what Bice calls a “risk profiler.” He set up a predictive model that allows them to alert account managers about which accounts may not be being served the best and determine what corrective action could be taken.
Business Analytics and Customer Segments
Bice also uncovered data that showed unexpected connections between customer segments and performance. “We did a deep dive for average monetary performance on accounts when we threw in the AdWords metrics.” Through this in-house data analysis, they discovered that fluctuations in revenue were much more useful in predicting cancellations in large clients than small ones.
Making New Business Decisions
Logical Position made company decisions based on their interpretations, making this a true business analytics project. The difference between small and large clients led the agency to change how they served smaller clients — improving onboarding would increase revenue.
Soft skills that Bice learned at GGU also come into play during the project: “At GGU, I learned how to research businesses and got a lot out of class discussions about how they work and use their data. Now, when I look at a data visualization with someone at work, they immediately say one thing, but I can give them an alternative explanation. I can offer an interpretation based on the business processes as a whole. That is very useful.”
On the strength of his customer-retention project, the agency created a new position for him in a new department called Operations Development that creates internal tools and processes to help the company scale efficiently. Bice passes on recommendations to programmers; and, with management support, they develop new applications.
The next step for Bice, he says, is to dig deeper into a data science role as his career progresses.