Share these talks and lectures with your colleagues
Invite colleaguesPractice paper
Time for marketing to embrace reinforcement learning
Journal of Digital & Social Media Marketing,
10
(2),
135-142
(2022)
Abstract
Since COVID-19 upended the world, marketers can no longer rely on historical data to inform their decisions. Channel splits have changed and online conversations have exploded. Marketing budgets have decreased as a percentage of revenue, meaning marketing funds must be used more effectively and efficiently than ever. Fortunately, the relatively new application of reinforcement learning — a data science approach — in marketing offers additional opportunities to gain competitive advantage using artificial intelligence. Unlike other types of machine learning, reinforcement learning uses algorithms that do not typically rely only on historical data sets, to learn to make predictions. Rather, these algorithms learn as humans often do, through trial and error, adjusting their ‘behaviour’ based on the outcomes of their actions. While the algorithms and computations behind reinforcement learning can be complex and sophisticated, its ability to deal with real-time decision making makes it an attractive option for marketers. This paper shows that with the right ‘business translator’ — that is, a person or team operating as the ‘glue’ between data science and business performance — sophisticated data science becomes accessible to commercial teams looking to drive performance improvements.
Keywords: reinforcement learning; data science; marketing analytics; change management; artificial intelligence; digital marketing
The full article is available to subscribers to the journal.
Author's Biography
Laura Murphy is the co-founder and Chief Executive Officer of Amplify Analytix. A former head of business transformation in international markets, Laura specialises in helping organisations embrace the benefits of data science to improve commercial performance.
Fernando Perales is Head of the Research Lab and Strategic Partnerships at JOT Internet Media, where he is responsible for the coordination of innovation projects, from developing in-house technology to participating in pan-European collaborations to test cutting-edge Big Data applications.
Anand Gopal is the Co-Founder and Chief Operating Officer of Voiro. An ex-operations research practitioner and Georgia Tech graduate, Anand is responsible for driving company growth, strategy and operations at Voiro.
Yordanka Gyurdieva is the Head of Product Strategy at Amplify Analytix and the product owner of SOLD! — Amplify’s award-winning contextual multi-armed bandit solution. She helps organisations translate business problems into hypotheses that can be treated with data science solutions, such as reinforcement learning.
Victor Gueorguiev is a senior data scientist and avid user of reinforcement learning. He uses his deep expertise in advanced analytics to develop, test and deploy data science solutions to solve customer needs and pains.
Pratyush Shandilya is a Data Scientist at Amplify Analytix. He has developed bespoke end-to-end data science solutions for clients in diverse industries. His expertise is in Deep learning, Reinforcement learning and Natural Language Processing. Pratyush holds a Master of Science from the Carnegie Mellon University.
