Written by: John Keenan, EVP/Marketing Analytics at Periscope.
Historically, inventory-related data ruled in retail, but more recently, Big Data has meant an explosion of different data types available to retailers, delivered at ever-increasing velocity and volume. Today’s challenges include how to harness this data to provide value beyond simple transaction analysis, and to utilize it before it becomes stale. The key to making transaction data actionable through analytics in today’s dynamic environment is understanding the context within which retailers are operating, and the context within which consumers are making purchase decisions.
Here are a few examples of how context might influence types of analysis a retailer might do:
- Set appropriate expectations for customer purchase behavior, and align with actual transaction history to drive personalized messaging and offers.
- Characteristics of your product or service, such as cost, complexity, degree of substitutability, need for variety, etc., often dictate average purchase cycles. Transaction history can identify how different consumers operate in your category and how that lines up against expectations, which in turn can inform a segmented messaging strategy intended to drive desired behavioral changes at the individual level: right message, right time, right channel.
- Be creative in how you use newly available data sources.
- Utilizing historical Point of Sale (POS) data from individual retail locations can help you understand what happened in the past, and forecast what will happen in the future, but those forecasts assume all influencing factors will remain constant. If you creatively incorporate factors like weather, changes in competitive presence, online search volumes, etc., you can get closer to why sales followed a certain trend, and ensure your marketing plans can flexibly respond as conditions change in the future.
- Analyzing the controllable and non-controllable factors.
- Sales are influenced by many environmental considerations, and by relating transaction data to the controllable and non-controllable factors surrounding each store location, effective local store marketing plans can be implemented. Outside of the store, this might include trade area demographic composition, drive times and traffic patterns, weather, parking availability, degree of competitive presence, etc. Inside the store, you can incorporate analysis of things like staffing levels, shelf placements and promotions, pricing and more.
Using different types of data to understand context, and relating it to performance metrics, is an important way to help optimize your marketing efforts.
