• by Laurie Sullivan  @lauriesullivan, July 13, 2015

Since marketers don't always understand how much weather patterns, compared with the advertisements, contribute to the sale of something now and in the future, Convertro, a division of AOL Platforms, plans to change. Monday the company will release an attribution model that considers 27 weather conditions.

Search marketers have been using weather data to target ad for about two years. Traditional media models account for the weather's impact on consumer action by calculating the average temperature during a week across all consumers in a region. Covertro will now tell digital marketers to what degree the weather could affect present and future purchase decisions on desktops, smartphones and tablets in specific regions by ZIP code across the United States.

Local weather, not just ads, influences purchases, said Amy Mitchell, general manager of Convertro, a division of AOL Platforms. "It's about planning and making predictions on how media will perform through machine learning," she said, adding that it's not necessarily in real-time, but within a couple of weeks.

The platform will tell marketers, for example, when too much snow in one week across specific ZIP codes in New York influence local residents to book a trip to warmer climates like Hawaii.

The platform will have the ability to build variables through machine learning to assign credit across TV, mobile, online, catalog, direct mail and other media.

Convertro will source and process anonymized user-level geographical data, time attributes, and daily and historical weather data such as temperature, precipitation from the National Oceanic and Atmospheric Administration (NOAA) for all ZIP codes.

Weather data will become the first step in several changes Mitchell expects as Convertro settles-in under the wing of parent company Verizon.