Most consumer packaged goods companies rely on analytics to assist them with understanding the health of their trade promotion. However, most also struggle to get from data availability to analytical insight because of the inability to overcome workflow issues surrounding existing analytics practices. This is particularly apparent concerning time-consuming and error-prone spreadsheet management. Companies can take control of their data management and turn it into key insights through adopting machine learning practices, thus ultimately allowing more informed, more profitable decision-making.
Before continuing, it’s important to address the misconception surrounding the term “machine learning.” Harvard Business Review’s article by Thomas H. Davenport, the president’s distinguished professor in management and information technology at Babson College, and cofounder of the International Institute for Analytics, “Move Your Analytics Operation from Artisanal to Autonomous,” defines machine learning as “any data-driven approach to explanations, classifications, and predictions that uses automation to construct a model.” In short, machine learning automates analytics.
Many are hesitant when faced with this term because they fear they will be replaced by a smart machine. However, while the technology is smart, it’s not the type of smart that allows it to operate on its own without human input and intervention. Rather, it does some of the heavy lifting, but still needs the human component to analyze what it compiles, turn it into valuable insights and execute the plan.
Trade promotion remains an area where the need to automate using machine learning is important because many CPG companies are only analyzing their top 2-5 customers and this only 1-2 times annually. As they automate the data compilation and integration process, they can analyze all of their customers as frequently as new data comes into the system. This means a more complete understanding of the business and the ability to identify where making individual adjustments for one customer will impact the overall organizational performance.
Trade Promotion Optimization takes this understanding to the next step by providing that elusive future visibility. Standard data management technologies, like TPM solutions, are effective at helping to reconcile events, but not equipped to tell you why one promotional tactic or strategy works better than another. TPO solutions, on the other hand, allow you to input specific constraints to define the optimal promotional mix to meet revenue, volume or profit goals for both the CPG manufacturer and retailer.
As a result, a TPO solution that can integrate with TPM is important, because doing so maximizes efficiency, eliminates redundant work and most importantly will result in a annuitized return on the annual trade spend. In addition, TPO should allow analysts to identify and correct data anomaly, overlay consumer and customer marketing events and adjust baseline frequency for seasonality and/or long-term trend analysis. With this, the role of the analyst transforms from being someone who spends most of their time compiling data to someone using their designated skills to actively interpret the data and provide valuable insights.
TPO eliminates the uncertainty left behind by a TPM solution's limitations with quantifying the ROI of promotions, thus enhancing future planning results. A TPO solution should apply the historical learning and the predictive models to make a recommendation with expected outcomes that can be brought to organizational executives to not only justify the significant annual investment.
Davenport states that “We can’t deal with it all using traditional, human-crafted analytical methods.” As data continues to grow and come in faster, the traditional means of handling it just doesn’t cut it. Having automated analytics allows you to successfully process your large amounts of data and spend more time working on finding the insights that will push your company forward to improved revenue management practices and sustainable growth.
What to read next: How Analytics Integration Brings Transparency