Marketers Are Using the Weather to Predict Buyer Behavior

Trends in tech and advertising have made the weather a highly sought-after data point, turning forecasters into predictive modelers of human behavior

“Temperature at the airport is 88 degrees, which is stupid, man, because I don’t know anybody who lives at the airport.”

– Al Sleet, Hippy Dippy Weather Man

Eric Danetz, AccuWeather’s newly installed chief revenue officer, picked up the phone and contemplated the climate. This was autumn in New York, yes, but with highs straddling the upper 60s, the afterglow of summer lingered like a lover’s fragrance, keeping the puffer vests and fiery foliage at bay. As always, Danetz took the weather in stride.

“The weather is always good at AccuWeather, my man,” he says. “Today is partly cloudy, and we’re gonna have some showers later today and early tomorrow, but otherwise not so bad. It’s going to be a potentially record-hot Sunday.”

The ease with which he rattles off the forecast is not innate. Danetz is new to this job, but he first used AccuWeather 27 years ago as a fresh-faced intern at ABC TV network’s flagship station in New York City. There, he crunched meteorological data for the station’s weatherman, Sam Champion, who would go on to report for “Good Morning America” and anchor shows on The Weather Channel.

Now, decades later, Danetz’s tenure as an AccuWeather employee spans less than four full seasons. But he’s already drawing on a long career in news to steer AccuWeather’s public-facing offerings toward a premium media outlet. He came to the company by way of Time Inc., CBS, McGraw-Hill and the Daily Beast. The new gig pairs him with another Daily Beast alum, Bill McGarry, and teams him with Deirdre Daly-Markowski, one-time executive director of digital at Hearst. That’s a lot of media moguls to report cloud conditions.

The staff starts to make sense when you look at the audience. AccuWeather claims a daily digital reach of nearly 2 billion users—put another way, as many users as Facebook. But whereas Facebook’s audience gathers on the central Facebook platform, Danetz says AccuWeather’s users are counted from a host of media channels that use AccuWeather.

“We provide forecasts for nearly every location in the world,” Danetz says. Users access AccuWeather on smartphones, tablets, desktops, TVs, radio stations and newspapers. They can even view AccuWeather’s own TV network on Verizon Fios. “If you go into Westfield Mall, or you’re downtown, here in Manhattan, you’ll see digital displays from us,” Danetz says. “If you’re in an elevator, there’s a company called Captivate [that broadcasts information in the car], and the weather on those screens is from AccuWeather. … It’s a tremendous footprint. I’m not going say we’re the size of Facebook in terms of branded usage, but in terms of people who are engaging with our content on a daily basis globally, it’s that big.”

Any content producer claiming to connect with billions of users is likely to draw attention from the ad market. AccuWeather is no exception. But digital displays in elevators and outdoor Manhattan only capture viewers’ attention for a few moments in transit. Danetz and others in AccuWeather’s newly assembled brain trust don’t just want people to check AccuWeather, they want them to experience AccuWeather. “We want to provide video. We want to provide immersive content so that they’re spending more time [with us].”

Part of AccuWeather’s strategy to cultivate brand loyalists is its enhanced app. It displays temperature, but so does the Apple widget that’s standard on every iPhone. Very sleek. Very sterile. Very Cupertino, California. “It’s basic for a reason,” Danetz says of Apple’s offering. “Apple wants it to be basic. They don’t want you there, they want you in the App Store.”

In contrast, AccuWeather’s app occupies a platform that’s closer to Facebook than Old Farmer’s Almanac. Users can toggle between current and extended forecasts, as well as real-time Doppler radar. More environmental conditions, such as humidity, wind speeds and precipitation accumulation are displayed by scrolling down past some ads. Further scrolling reveals allergen information of particular interest to health care marketers, who can advertise their health solutions next to high-pollen alerts. “Health care is a major client for us, whether it’s a Claritin, or other allergy-related medication,” Danetz says.

Beneath the allergy information are news video packages created in house by AccuWeather, mostly reporting footage from severe weather or natural disasters, such as West Coast wildfires or Indonesian volcanic activity. All play after a preroll word from sponsors. Beneath those is the extended local forecast, another programmatic ad and, at the very bottom of the page, astrological data about the sun and moon.

A menu button on the lower right-hand side lets users report their local conditions, upload their own videos capturing weather events and share AccuWeather content to external social media networks. For a certain segment of the population fixated on the weather, this is the place to be.

“Weather is an important topic in terms of content, but it’s also a social one,” Danetz says. “I’ve run news organizations where dwell time is long. The reason you’re going to spend time on weather, whether it’s a site or app platform, is because you want to know what’s happening. You want to understand how to plan for your day. Or if you’re a business, [you want] to understand how your business is going to run predicated on the weather for that day.”

Atmospheric Data

The app is only the beginning of AccuWeather’s beefed up offerings. In a way, it’s merely a lure, used to build the currency it and other weather organizations really traffic in: data. Every user that logs in to the AccuWeather app from a smartphone gives the company their location. Like other apps, this allows weather media companies to build profiles and sell targeted ads based on geography. Unlike other apps, the extensive weather measurement network means AccuWeather and its competitors can merge location data with environmental data to create a customized ad experience. Think umbrellas during a rainstorm, or sunscreen throughout summer.

Location data is particularly hot right now. The Location Based Marketing Association finds in its “2017 Global Location Trends Report” that more than half of all companies say they use location data to target customers. A research paper published by MDG Advertising found 72% of consumers report engaging with calls to action if they receive them while near the physical location associated with the message.

Users of the AccuWeather app understand the necessity of sharing their location data. Weather apps only make sense if their information is geographically relevant. Weather and location are so inextricably linked that it makes sense for AccuWeather to be in malls and elevators because conditions there can be different from the home, office or immediate travel destination of users. Weather, location and transit are so interconnected that Danetz is exploring options to integrate AccuWeather information on top of online maps.

“When you go on Waze or Google Maps, all it tells you is traffic information now and how long it’s going to take you to get to your location,” Danetz says. “What if there were weather information there? It’s one thing to wake up in the morning and check the weather to see what to wear, what type of personal products to use, but we’re serving over 32 billion data requests a day.”

For all the attention AI and machine learning have received, Danetz says AccuWeather has been rooted in data science for 55 years because it’s key to understanding heavy media investments being made by AccuWeather and others. One competitor has coupled media and data solutions with AI engagement to form uniquely powerful personalized advertisements.

Weather Channeled

The weather was even warmer the following week as Carrie Seifer, vice president and chief revenue officer of Watson Advertising (formerly The Weather Company), traversed the streets of New York en route to a midday rendezvous. “It’s beautiful in New York,” she said by phone.

Six months into her position, Seifer is inundated with requests to tailor her weather data—along with the rest of the Watson product suite—into help for marketers seeking smart, personalized, weather-driven campaigns. “You can imagine weather data is only limited by your creativity. It affects so many decisions made by people and businesses,” she says.

Two years ago, The Weather Company was sold to IBM. The Weather Channel TV network, which it owned, was not part of that sale. The TV channel remains in the hands of a consortium of owners, including NBC Universal, The Blackstone Group and Bain Capital. The new owners struck a licensing agreement with IBM to use The Weather Company’s forecast data and analytics on The Weather Channel platforms.

In acquiring The Weather Company, IBM bought a host of technological assets capable of monitoring environmental data points across the globe while serving three distinct segments: consumers, B-to-B and marketers. Adding more confusion into this flurry of activity, the marketers segment is now called IBM Watson Advertising. The rechristening reflects not just which product gets top billing but how far of a leap weather data has made in the realm of marketing. Weather is now seen as another key data point in Watson’s vast ecosystem. But more than that, marketers can now come to weather media companies with extremely sophisticated asks.

During research for a Pantene campaign, the Leo Burnett agency discovered that women often switched hair products on bad hair days. Armed with that insight, the agency turned to Watson to push out smartphone alerts warning consumers about imminent bad-hair weather. Consumers were urged to avoid a coiffure catastrophe by heading to their neighborhood Walgreens (Pantene’s partner on the campaign) for some Pantene preparedness. The result: a 24% increase in sales of Pantene products at Walgreens stores in July and August of 2013 compared to the same two-month period in 2012.

Yet, even this adroit example eschews mention of the most cutting-edge services offered by The Weather Company. Once, marketers could use weather companies as one of many media suppliers (a space for their current ads) or as a data supplier (information for their future ads). Now a third leg is propping up a table of targeted ads: AI.

“I call it DCO [dynamic creative optimization] on steroids,” Seifer says. “Imagine if you could merge Apple’s AI technology to DCO, so that it wasn’t so manual to have a sophisticated conversation in the ad unit,” Seifer says.

“We were bought by IBM and now have access to a lot of their product offerings for marketers, so for us to only bring publishing solutions to the marketer was limiting,” she continues. “Media, data and technology are what make up Watson Advertising. That weather product that you know and love? That’s in our media channel. That data that you wish you could apply to your entire buy? That’s our data channel. And do you want to use AI in your business, so you don’t have to hire 4,000 data scientists? That’s our AI channel. That’s our technology channel.”

Picture opening your phone to check the weather only to encounter an advertisement that prompted you to ask questions about the product it featured. Without leaving the app or involving customer service reps, the advertisement can respond to specific queries you have about the product: What color scheme does it come in? How does it improve on past versions? How does it compare to a similar product?

Such an ad is not only theoretical, a case study already exists. Toyota blasted ads for its Prius Prime throughout The Weather Channel App and weather.com for 12 weeks last year. Billed as an industry first, the ads allowed consumers to have personalized conversations with AI software that informed them in a self-contained, robust experience.

“We were able to use Watson as a [chatbot] that went between our potential customer and the products that they’re trying to talk about,” says Chris Pierantozzi, the executive creative director at Saatchi & Saatchi who spearheaded that campaign on behalf of Toyota.

Through the AI component of Watson’s content management system, consumers could obtain facts about the vehicle that would not typically be featured in a display ad. This information served as a data set in its own right because it informed Toyota and Saatchi & Saatchi on which questions were most likely to be posed by prospective buyers.

“We would have never thought that people would have cared about how fast the Prius Prime can go, but we actually had quite a bit of questions around speed,” says Lily Won, communications director at Saatchi & Saatchi. A common question was how quickly the Prius Prime could accelerate from zero to 60 miles per hour. Also popular were questions about self-driving capability and servicing the vehicle.

The most frequently asked questions turned out to be highly technical. Users who interfaced with the ad wanted to know more about the Prius Prime’s dual-motor drive system and remote climate control. The ads targeted those who were in the market to buy a new vehicle as well as auto-tech enthusiasts, which might also explain some of the wonky questions, Won says.

The campaign performed about three times higher than the benchmark it was measured against and showed positive lift in key brand health metrics. While there wasn’t a specific set of weather-related factors that influenced where the ad was pushed, some of the analytics indicated weather conditions impacted effectiveness. People in clear-weather conditions—perhaps more readily able to imagine themselves coasting down the freeway on a sunny day—were more likely to engage than people in a rainstorm.

“The quality of the emotional state is plainly influenced by the weather states.”

– Edwin Grant Dexter

“Weather influences: an empirical study of the mental and physiological effects of definite meteorological conditions” (1904)

More Than a Feeling

The warmth left New York as October closed. By early November, the five boroughs were like one large vegetable crisper. Fortunately for Serge Matta, president of location-based marketing and advertising platform GroundTruth, he was working off-site in considerably warmer Mountain View, California.

The former CEO of media measurement firm comScore, Matta ascended to the helm of GroundTruth roughly 15 months ago. Around the same time Matta was being onboarded, GroundTruth, then called xAd, purchased the popular weather app WeatherBug. Since then, the company rebranded to reflect its evolution from media supplier to data supplier. A large part of the new iteration specializes in research showing how weather influences consumer behavior.

The impact of weather on human activity has a long history dating back at least to 1904, when Edwin Grant Dexter, an early biometeorologist, published extensively on the subject. Recent GroundTruth reports first published on The Drum found Americans are happiest when the temperature is between 61 and 75. Months with this type of weather coincide with a 15% increase in retail foot traffic in California. Months with temps above or below this range see a 10% drop. Californians are also 32% more likely to head to a movie theater on rainy days. The same can’t be said of Seattle, where rain is 46% less likely to influence traffic than in the Golden State.

WeatherBug is not alone in exploring the effects of weather on buying patterns. Information provided by The Weather Company concludes that juice sales increase during periods of high winds and low temperatures. Soda sales tick up during snowy or rainy winter weather, liquid detergent sales increase during periods of above-average precipitation and below-average temperatures, and tire sales spike following a string of clear-sky days with above-average temperature.

The weather’s influence on consumption doesn’t only manifest in purchase behavior. Last February, AccuWeather partnered with Spotify to launch Climatune, a music microsite that features playlists of popular song selections in various cities during certain types of weather. The results are based on data from 85 billion anonymized Spotify streams in more than 900 cities from November 2015 to November 2016. The populace of Perth, Australia, for instance, is partial to Gorillaz’s “Feel Good Inc” on sunny days, while clear nights in Chicago mean many locals are queuing up Seal’s “Kiss From a Rose.” Overall, sunny days inspire more positive, high-energy song selections, and rain produces subdued, sad-sounding tracks that are more likely to be acoustic than electronic. 

“Weather data predicts emotion, which directly correlates to consumer behavior at a point of need,” Danetz says. One client, he notes, uses AccuWeather-provided analytics to chart the seasonal change in preference for iced or hot coffee.

All weather companies traffic in these insights, but perhaps none lean on them as much as GroundTruth. Shortly after the acquisition of WeatherBug, the company rolled out Weather Triggering, which purports to enable marketing to target against many weather conditions.

“Anyone can sell a coat in the winter or an umbrella when it rains, but how will a week of high humidity or air pressure impact foot traffic and sales?” Matta asks, quoting GroundTruth founder Dipanshu “D” Sharma. “Our acquisition of WeatherBug and launch of products like advanced Weather Triggering are enabling us to bridge that gap for our partners by combining location and weather data to help brands anticipate and adjust to weather patterns that are less-understood.”

The service, which launched in August, relies on environmental data extracted from the WeatherBug network coupled with biometeorology to pinpoint ideal conditions for brands to blast location-based promotions.

Attitudes can differ widely from one region to the next, so Matta believes marketers should account for seasonality, the increase in weather uncertainty and regional, weather-related foot-traffic trends. GroundTruth has broken these insights down by core verticals and by states.

“Record-low temperatures in New York and New England in early March led to a 27% decrease in [quick-service restaurant] foot traffic,” Matta says. “Interestingly, fast-casual restaurants saw the smallest decrease as consumers wanted to spend more time inside. For the retail category, visits to major toy retailers are 4% higher than average on high-precipitation days. This tells us that brands should be adopting a local marketing strategy that’s customized for their target consumers by region or state. This ensures they’re speaking to their audience’s current behavior and needs.”

Since the service is so new, there’s no telling how much of a difference the weather itself can make for particular brands. But climate has been a prime driver of human actions for thousands of years, and as forecast models predict more extreme meteorological events, it’s a safe bet that consumers will continue to be responsive to outdoor conditions.