Google BigQuery Helps Show How Weather Influences Shopping

Interactions, a marketing firm for supermarket chains and big box stores, announced some interesting findings based on analysis with Google’s BigQuery and Tableau, measuring the effects of major...
Google BigQuery Helps Show How Weather Influences Shopping
Written by Chris Crum
  • Interactions, a marketing firm for supermarket chains and big box stores, announced some interesting findings based on analysis with Google’s BigQuery and Tableau, measuring the effects of major weather events on sales.

    “This case study is just the first analysis in an ongoing relationship between Interactions, Google and Tableau to provide retailers and CPGS with data management solutions and insightful analysis,” a spokesperson for Interactions tells WebProNews. “While the results of the initial study may seem obvious, this type of analysis provides retailers with specific detailed information not previously available that will help them launch more successful marketing efforts resulting in increased basket size and cannibalization of competitors’ shoppers.”

    Using BigQuery and Tableau’s visual analytics software, Interactions layered “fine-grain transactional-level sales data” with multi-source, detailed regional weather data, and is using it as a predictive tool during and prior to specific weather events in order to help retailers minimize (or eliminate) out-of-stock issues, optimize item assortments in high-demand categories, and increase sales in those categories.

    “Early results suggest that these insights could deliver significant benefits, including boosting the return on investment for marketing and advertising spend and improving targeted and individualized shopper communications resulting in increased basket size and cannibalization of competitors’ shoppers,” the company says.

    The analysis specifically focused on identifying negative weather events, classifying them by severity, and measuring the effects they had on sales before, during and after those events. The company says it was able to track new patterns in sales and shopper behavior.

    “Interactive performance of Google BigQuery, combined with Tableau’s intuitive visualization tools enabled our analysts to interactively explore huge quantities of data – hundreds of millions of rows – with incredible efficiency. In some cases taking analysis that would ordinarily require a week down to just hours and minutes. This time-to-insight was previously impossible,” said Giovanni DeMeo, Vice President of Global Marketing and Analytics for Interactions. “It enabled us to visually share that information with our retailer and CPG partners, and use it to enhance in-store activity and increase sales. This is only one of an infinite number of ways that we will now be using big data to improve the revenue and profitability of our partners.”

    “Retailers have access to massive amounts of complex data to help them make good decisions. The trick is to find a way to easily visualize and analyze it effectively,” said Francois Ajenstat, Director of Product Management at Tableau. “By combining the flexibility and horsepower of Google BigQuery with Tableau’s visual analytics, Interactions has delivered insights that were not previously revealed. Retailers and CPGs will now be able to make real-time data driven decisions to inform their business.”

    According to Interactions, it was able to reveal, down to the product level, which items had the most significant change in sales (both increases and decreases), and what varied in shopper behavior for similar weather events, considering the time of day, day of week, geographic location, and proximity to competitor locations. The data was actually able to identify 28 categories with significant changes in sales (compared to the control).

    Based on the data, one day before statistically similar weather events, sales in those categories spiked from 20% to 261% over the same day in the previous year, and dropped in sales during the peak of the event and for four days after. This happened not only in regions that actually experienced the event, but also in those where the event had been predicted, but never occurred.

    “If weather reports predicted a storm a week ahead, people still waited until the day before the event to do their event-specific shopping,” says Interactions. “In one scenario, and contrary to every other shopping behavior, this resulted in a huge spike in Monday sales over the preceding weekend for a predicted Tuesday weather event.”

    Interactions, Google and Tableau are discussing more details about their findings at the Tableau European Customer Conference in London on Tuesday.

    Get the WebProNews newsletter delivered to your inbox

    Get the free daily newsletter read by decision makers

    Subscribe
    Advertise with Us

    Ready to get started?

    Get our media kit