At the end of one year and the start of another, listicles of the top trends to watch crop up across industries. Allure predicts the makeup that will dominate TikTok tutorials, Pantone highlights the colors that will dictate design decisions and Architectural Digest foretells the cabinet choices we are apt to make in our homes.
When it comes to software advances, the tech industry is no different. Forbes, Deloitte, CNN and CB Insights have published reports and articles about 2023 tech trends, like data governance, the metaverse, the Internet of Things and artificial intelligence.
At Camelot Illinois, Richard Bowman has his eye on one trend in particular — machine learning. He’s expecting it to shape how lottery operators like Camelot engage with customers.
“Lottery players increasingly expect experiences tailored to their wants at a given moment,” Bowman said.
Though machine learning already influences how the lottery industry approaches that challenge, Bowman sees its role expanding to the point where machine learning models are a cornerstone of successful customer engagement.
Camelot has already developed in-house machine learning models that analyze customer interactions and generate actionable insights for the company. And last year, Camelot tested promotional campaigns powered by machine learning.
They found that a model-based approach outperformed alternatives. As a result, Camelot is significantly increasing its investment in promotions as a result of its findings.
“We now know that machine learning can deliver incremental sales,” Bowman said. “This year, we’ll scale it in other high-opportunity areas, like CRM campaigns, product analytics and automated analysis of business performance.”
Built In Chicago sat down with Bowman to find out how his team is adapting its roadmap to expand upon machine learning and exploring new tech like it.
Camelot Illinois operates the Illinois Lottery. At Camelot Illinois’ core are consumers and social responsibility. Camelot funds schools, capital projects and other causes.
How is your team’s roadmap adapting to that trend?
Our data team will expand their ML skills while simultaneously investing in the technology needed to support this effort. As the team works toward those objectives, we will continue to collaborate with stakeholders across our business to identify additional opportunities where data, analytics and ML can drive sales and business efficiencies.
From a tech perspective, our in-house data platform is cloud-based and capable of delivering ML scores and outputs in real time. We will upgrade a handful of our processes this year, including moving to a largely serverless architecture. This will ensure we are well positioned as the volume of data, analytics and ML services we provide to the business grows. It will also enable us to get more output from our existing data budget.
From an ML perspective, there are three goals: work collaboratively with our marketing team to ensure that our in-market models continue to perform well; deploy additional promotion powering models; and extend the scope of our ML activities to cover CRM campaigns, product analytics and the automated analysis of in-store retail performance.
How does Camelot Illinois empower and encourage data scientists to explore and learn new technologies?
Most of our work is based on briefs created in collaboration with business stakeholders. However, Camelot cultivates a culture of learning by failing fast. We encourage our data team to undertake proof of concept work. In practice, this means developing new skills to find solutions to a specific business problem.
Camelot has made several significant steps forward thanks to this approach. About a year ago, our Data Science Manager Saurabh Pal was encouraged to work on a proof of concept ML model that could power customer promotions. We both believed this was a high-potential opportunity for ML, but understandably, business stakeholders required evidence that this could increase sales.
In the end, we developed a plan to power 80 percent of our promotions using ML models from proof of concept. Our CFO and VP of marketing are confident enough to significantly increase the investment we make in promotions. As part of this process, Saurabh enhanced his ML knowledge. He mastered the XG Boost algorithm and developed his AWS skills in deploying models live into production.
