Brian Hull is a creative leader with over 15 years experience in driving award-winning solutions for some of the world’s largest advertising agencies, marketers and brands.

With executive creative posts at New York’s Grey, Organic, and G2, Brian has amassed a variety of B2B and B2C category experience across multiple marketing channels.

Be it Times Square signage, to in-store, in-dash, on-pack, on-line, mobile and more, Brian is a strong believer in creating unified marketing experiences that focus on being relevant, engaging, intuitive and beautiful.

Newly acquired by IBM, Brian leads the The Weather Company's Global Creative Lab teams utilizing Watson’s cognitive capabilities along with the ultimate trifecta of data–weather, location and time–to concept, create and deliver advertising products, services and experiences across all Weather Company properties.

Follow Brian on Twitter


Episode Transcript

J. Cornelius: Hey everyone, thanks for joining us. Today we are excited to have Brian Hull. He is the head of Global Creative Labs for The Weather Company, which is an IBM business. Hey Brian, how are you doing today?

Brian Hull: Really good J, thanks for having me on.

J: Thanks for joining us. Just as a way to get started, tell people a little bit about your background, and what brought you to The Weather Company and what you're doing there.

Brian: Sure. I started off as a designer many moons ago, a passionate and creative world, a traditional designer who got turned onto computers back in the 80s and 90s. I always swore I would never become a computer artist, I was always into fine art, but that attention to detail got me into animation, and animation got me into electronics and computers and old things like Softimage and all these old animation programs that were really hot back in the day.

J: That seems to be a common path for people.

Brian: Yeah. I had to go back to school to learn this stuff, all these programs, and I remember, I think I spent about three weeks creating a globe in a 3D program. It took me about three weeks to map out the globe, and then about another 30 hours to make it turn about a quarter of an inch. My professor at the time said, "Brian, there's this new tool that just came out called Macromedia Flash you might be interested in." Within probably three days I was able to do what I couldn't accomplish in 30 days, and the other side benefit on top of that was, I was able to publish it out to this new thing called the World Wide Web, and have my friends in Australia and all over the world comment on it and play with it and share it. That's where the bug bit me.

In that front end, the passion from front end, I started to realize if I need the front end to sing I'd better learn the back end, so I got into coding so I could make those two things sing together. That kind of got me into the world we're in now, evolving quickly from designer to coder to then a UX designer, into now a strategic product creator. With The Weather Company, I had a lot of big background within advertising agencies and digital, and cutting my teeth there in San Francisco and New York, but The Weather Company really enticed me with what they were doing with data, especially within location. What I like to call the ultimate trifecta of data, which is weather, location, and time of day. It affects every single thing that you do. What you eat, what you drink, what you wear, where you go on vacation, the car you drive. Then they lifted the hood on a lot of the data with me and showed it to me, and that's when I went okay, I've got to get onboard. Since we've been acquired by IBM, and now I'm working with a lot of those type of weather data driven experiences from advertising new product, and now weaving in cognition and Watson experiences.

J: Yeah, I think a lot of people probably underestimate just how much the weather impacts their daily decision-making. Is that the kind of stuff you're looking at?

Brian: Absolutely. We're moving away, we're no longer a weather company, we're a data company. You hit the nail right on the head. Weather really does severely impact ... When you think about it you can start connecting the dots very, very quickly. From your medical, your health and wellness, your emotional well-being. If it's constantly gloomy or if it's sunny, if you're an allergy sufferer, your mods are severely affected, and the regions that you're in. In the science and the data behind that, we've been able to articulate and understand and prove out that, depending on cloud coverage, wind speed, and dew point, that yogurt sales increase. Who'd've thunk.

J: That's fascinating.

Brian: Similar scenario can happen for laundry detergent, or carbonated beverages. Let alone automotive car sales. It affects everything. We're really pivoting more and more towards becoming a decision platform businesses.

J: How do you understand that data? We talk about doing user research all the time, and that you really need to understand users' behaviors and incentives, and what tasks they're trying to accomplish. We talk about that stuff a lot. How do you gather that information and then use that to drive some of the decisions that you're making around product and roadmapping features and that kind of thing?

Brian: Sure. There's a couple of ways of doing it. One is, location's a big one. It's not a big brother type of scenario here where we're tracking people. 85% of our users actually opt in to share their location with us. The numbers dwarf Google and Facebook combined in terms of allowing people track their location. They realize that they have to let us track their location in order to give them life planning, and many times life saving, weather alerts. How can I give you a weather alert if there's a lightning nearby, or a tropical surge is coming your way, if I don't know where you're at?

J: Sure, or even snowstorms and flash floods and all kinds of stuff.

Brian: You got it. We're able to use that location to understand patterns and behavior, and though we're not able to get too forensic, we don't know your name, or down to device, ID or anything like that, but we can tell, are people going in near car repair shops often? Are they going to fast food restaurants? Are they nearby shopping malls? That person's probably an auto intender. You might want to start serving them advertising elements that are products, services, and experiences that are relevant to where they are in that part of their life, looking at their portraits and their rituals. That's why our advertising is so, so effective. We're not carpet bomb with irrelevant programmatic scenarios. We're very, very targeted, very, very smart, in understanding behavior and profile data, very high-end behavioral profile data, so that what we're offering ... Again, we're using weather, location, and time of day. I'm not going to offer you up a swimsuit if I know it's 20 degrees outside and with a high wind chill factor. It's very, very relevant and very, very engaging.

J: What have been some of the more surprising things that you've discovered as you've dug into some of those behavioral information, and how that dictates what people are doing?

Brian: It's an interesting question. It goes back to the forensic nature of that ... I didn't finish answering your first question really, it's a nice segue, is that when you understand the matrix of possibilities that are tied to weather conditions and reactions, it's quite fascinating. A lot of things make sense right off the top of your head. It's cold; show an ad for a sweater, or a North Face parka jacket, or a Patagonia, et cetera. You're in the northeast in the winter, I'm going to present you with an experience that maybe shows a car that's four-wheel drive versus a convertible. That stuff's easy-breezy. That's simple block-and-tackle kind of analysis. But getting into understanding how people react to dew point, or wind speed, or cloud coverage, these things are absolutely fascinating. Our brands and advertisers and our partners that we work with find them absolutely fascinating as well, and they're able to make pretty amazing decisions that influence not only the consumer side, but also the business side.

We power a lot of communications for insurance companies. We're able to forecast, not only behavior, but we're able to forecast the weather obviously and say, "Hail storm's coming. Email all of your policy holders and tell them to get inside of a garage if they have one. You're going to save millions of dollars in car damage claims." We're also able to forecast behavior for law enforcement companies globally by mapping lunar cycles, and then tying those weather conditions into lunar cycles. We've all heard that when it's a full moon people go bonkers; well it's not just full moons, there are other type of lunar cycles combined with lunar cycles that also affect behavior in people. That kind of stuff's just absolutely fascinating.

J: Yeah, it's really fascinating, because I'm thinking the common wives' tale is that people go crazy under a full moon, so what other situations have you identified where behavior is different than what you might expect, or where there's an anomaly in the way a group of people might behave based on weather and the other factors that you're tracking?

Brian: Sure. I can't go too deep into it because probably a poison dart would come out of the ceiling and pop me in the neck, especially when it comes to that behavior. Or especially when we're working with law enforcement or whatnot, or military. But there are, as I mentioned earlier, how does dew point and wind speed affect yogurt consumption, or increase purchase behavior of laundry detergent? These types of things, that math and that science that goes into that, is still being figured out as we speak. A lot of times we can't quantify and qualify it, we just know it to be true based upon the history of weather patterns, and then also performance data. Whether that performance data can be sales from a dealership, or sales from dealers nation-wide, when we start mapping patterns of weather over let's say a three to five year period of time, and if we have a three to five year sales information, we can start making some pretty interesting correlations. There's chaos theory behind all that stuff. Weather is a finite science. That's a pretty solid way of understanding behavior and mapping behavior, but maybe not understanding exactly why wind speed, dew point, and cloud cover makes people behave that way.

J: That's pretty interesting. As I'm sitting here listening thinking about predicting behavior, some people may have their doubts, because as everyone knows, the weather man is wrong 50% of the time. Specifically in some parts of the country where the weather can change pretty quickly, how do you get confidence in the predictions? If the weather changes, how does that change the predictions that you're going to make around people's behavior?

Brian: Great point. We move quickly. We have to. This is a 24-7, 365 operation, and we pivot very, very quickly, because you hit the nail on the head. Weather isn't always consistent. That's changing and evolving literally as we speak. Things that were never heard of or ever experienced in centuries are now being experienced. With hurricane Sandy and others things are constantly in flux, so we're always expecting the unexpected. Pivoting to those scenarios, there's a lot of if and else scenarios that we plan for, so that we do pivot we're as relevant and as accurate as we possibly can be. I deal more on the communication side and the advertising and marketing side of utilizing that weather data, and now cognitive data, to interact with people. Not so much on the weather product side. But I can tell that I'm proud to say that The Weather Company has been voted and found to be the most accurate weather forecast multiple years in a grow globally, and that we're constantly evolve our infrastructure to make sure that we continue to make the most relevant and the most accurate weather forecast on the planet.

J: Because ultimately that's what people go to the weather channel for, is they want to make sure that it's not going to rain today, or that maybe they want to know if it is going to rain today so they don't have to water their garden. Another misconception that I know I've had in the past is that all the weather predictions come from generally the same source, and that The Weather Company may be one of those de facto sources, so all the apps are the same, or all of the weather predictive sources are the same, when that's just really not true at all. How do you differentiate yourself in terms of prediction? Is it really just letting people figure out that you're more accurate, or is there something else involved?

Brian: That's a good question. Couple of things. One, we actually power, Weather Company data, actually powers most of the other apps and the weather experiences that are out there, with the exception of a few competitors. We power Microsoft, we power Google, we power Facebook, et cetera. Most of the ones that people use at scale globally. But where we really shine is, there's a heritage and a trust that's been built up that's really owing to the cable side. We are the Campbell's Soup of weather. We are the trusted heritage brand. We're not new on the scene. We've been around for quite a long time.

J: Yeah, going way back to early cable. There was two channels that were on every box: the guide that told you what was on, and the weather channel, which was always channel three or something, so you always knew where to go.

Brian: Yep. Over that time though, we've been able to learn and understand, what do people need? What are the elements that they're looking for? Also that varies region by region and country by country, and locale by locale. Some people want boating information and some people want skiing information, easy at their fingertips, more easily accessible than others. We're introducing a new cognitive onscreen with our new mobile app experiences, that start dynamically presenting data that is more relevant to you.

That's another thing that's interesting too, with not only the content of the weather products but also the advertising. That even extended to television. Because when IBM acquired us they actually did a split of ... IBM doesn't normally own or acquire consumer brands, especially ones that are like us, at such a high scale. They separated out the cable side from the data side, from the mobile and the web side. They did that because they wanted the data, really, to power product services and experiences they could offer their clients all the way up to enterprise level organizations.

Because that weather data comes in handy, not only for a lot of IBM's large customers, but for Watson as well. So we're feeding in Watson to start helping enterprise level brands make decisions, and also helping consumers make decisions. That goes all the way up to some of the largest enterprise level organizations that you can think of, globally to you, J. If you download that app you're going to have a very different experience than your next door neighbor depending on how you use it over time. It's all about personalization, it's all about relevance, so that those recommendations are meaningful to you. It's not just about, to your point, is it sunny or is it going to rain. You can look out your window. It's about forecasting and planning your life.

J: It's funny, because I use the example of a weather app sometimes when I'm talking with startups around understanding people's behavior in their native habitat. If you go to the zoo, seeing a monkey just sit in a cage isn't very entertaining, it's not good for the monkey, but if you can go to the wild and see the monkeys swinging around in their natural habitat, it's a much more authentic observation of what they do. Using a weather app, a lot of people assume that you're going to be in the same environment every time you're looking at the weather app, like you're always sitting at the breakfast table, or you're always sitting at your desk, or something similar. I tell people that's not true, like a lot of times you're going to be sitting on the side of your bed putting your socks on, or you're going to be looking at it right as you get out of the shower, or whatever it might be, and we never know. How are you looking at different user behavior, and using that information to guide that product experience?

Brian: That's a great question. I didn't know where you were going with the monkey example in the zoo, but that all came together. Really, right now it's in its nascent stages of behavior, and access to data is ... The paradigm is shifting. Because it used to be news and cable and now it's mobile and web. Now we're seeing more and more it's about voice, UX or VUX experiences. Currently it takes me five taps on my mobile phone, a minimum of five taps, to order an Uber car. Or I could just tap it once and say, "Siri, I need an Uber downstairs to take me to JFK in 20 minutes to the Delta terminal." Done. A year ago did you think you'd be talking to a smart speaker in your living room? Let alone-

J: No. We thought only Iron Man could do that.

Brian: Exactly. Start thinking about autonomous vehicles and that captive audience. What are people going to start doing when they're not driving? They're going to start looking at screens and they're going to start talking and using voice, or using kind of narration experiences. You can gain some information about where I am, where I'm driving by, or where are restaurants that I may like. By the way, my son has a peanut allergy, so make sure that those restaurants cater to my son with a peanut allergy and my wife that's a vegetarian. All these type of personalized experiences that are delivered to consumers across multiple touchpoints. It's now, the ecosystem's hard to nail down. Screens are important, and they're kind of a good paradigm right now, but those screens are changing. Voices seem to be a big part of where we're going. Of course mobile is not going to go away any time soon, but how we interact with our mobile devices and the pieces of glass around us are definitely changing.

J: Yeah, solely. Even in my home, I think we've shifted from looking at a screen to get weather to just asking Alexa for it. Sometimes those predictions are usable if I'm staying near my home, but if I'm staying somewhere else I might not know exactly the zip code or the place that I'm going to that's going to help me understand if it's going to be clear or cloudy or rainy or whatever. I think there's still a little bit of work to be done there. I'm just curious, how are you gathering that feedback? You said you're in the creative labs; are you bringing people in, are you doing field studies? What kind of work are you doing that's getting that information for you?

Brian: All the above. It's all out the couple of ... That goes back to our design thinking principles. With IBM they have three things that really drive them, which are a focus on user outcomes, restless reinventions, and diverse empowered teams working on all these. The first is ... They're all very, very important, but the first is to really start focusing on, why? Not only are we focusing on the outcome, but how do we get people to start? How do we get people to interact with these experiences so they can understand the value proposition, so they can understand the reward mechanism, whether it's entertainment device, information, sales offers, a voice, so much more. That goes down to really basic ethnographic research, like an archaeologist understanding, what's the environment that people are in? How can they access information that's valuable and relevant to them? Whether it's in their car, their office, their home, their school, their gym, and their supermarket, wherever they may be?

The restless reinvention part though, that's where things get interesting. We try and stay essential by treating everything as a prototype. Nothing's ever a set it and forget it. I'm finding that web experiences are starting to get, not stale, but they're in boxes, and there's deep, deep science that goes into user experience. You remember how everything used to be above the fold, and that was a standard and a rule that existed? Now that paradigm has been broken. Now it's boxes, and commerce has taken over in different ways than before, especially in mobile experiences, and then geospecific and location. Making those recommendations, like hey, not only do we know what you like, but we know what's around the corner. Here's a map and here's a coupon. Have a nice day. Starting to get really, really hyper local with the messaging, the offers and incentives, and the information that we're giving to you, whether it's the weather or whether it's a coupon for a dress that's on sale.

It's all about constantly torture testing at every step of the game, whether it's in a vehicle, in store, on a shelf blade, on a product. Whatever it may be, consumer or a business ecosystem, constantly testing, and there's never a set-it and forget-it. Especially where we're at right now, it's too nascent. Everything's brand new, and one of the lines I use over and over again is, we're building the bridge as we're driving across it, constantly.

J: Yeah, or assembling the plane on the way down after you jumped off the ground.

Brian: Hopefully not crashing down, yeah, you got it. Staying up.

J: Yeah, exactly. Are you concerned at all about people's perception of that type of hyper local and behavioral based ad targeting? Are you concerned about that being problematic?

Brian: If it's disruptive, if it's not relevant, you bet, absolutely. If it adds to the experience, if it's opt-in, that is where ... It's all about relevance and a value exchange. If I'm not getting people offers, incentives, services, products, and experiences that are relevant and empowering them, helping them make decisions or propelling their world and their life, then yeah, that's disruptive, in a bad way. In a bad way. That's where ad blockers come into play, and that's where people are becoming ... Heck, even looking at study and research over a dozen years ago, people are really good at teaming up digital advertising experiences. Really, really good. They are trained like an athlete to hop over the hurdles of disruptive ad experiences so they can continue to consume editorial content or the other content they originally came to this experience for. Ads overall can be quite challenging in terms of hyperlocal, or targeting you in too much of an invasive personalized way. That's something that we don't participate in. I don't see anybody being successful in doing something like that. That's turning your ad into a robo call.

J: I think we've all been to those pages that are just littered with ads, and it's like playing a game of find the content, where you're trying to scroll through and actually discern, what is an ad, what is the content that I actually came here for. People are getting really, really well-trained in tuning out advertising that they don't deem as relevant.

Brian: Absolutely, absolutely. It's all about that personalization, and the relevance, in order for that to be engaged upon and in order for it to be rewarding for either the brand or the advertiser. We see ads in a very, very ... The shelf life isn't much longer for typical ad experiences as they are right now in boxes and rectangles, and what we used to call barkers or shoutouts, constantly barking at you to get your attention.

J: Yeah, the old banner ads are dying.

Brian: Yep. It's turning more into what we think of as a cognitive assistant. Something that's helpful that I can interact with or engage with, whether it's with my car or with my appliance or with my mobile device, or on our applications, our product services and experiences, no matter where we are, that we can engage with them and start having a relevant rewarding kind of utility experience. Again, that's what Weather really is at the heart of, we're helping you make decisions.

J: So it's shifting more from being a, "Here's a thing you might like," into a recommendation that's actually tailored to your specific lifestyle and behavioral choices.

Brian: You got it, and we're getting more and more to that type. Combining that with the input, this kind of goes back to your other question: How do we tweak it? We have feedback loops constantly. Was this helpful, was this not helpful? Again, going back to that treating everything as a prototype. Also utilizing multiple, not only teams, but also diverse people. Not constantly marketing and testing with millennials, we're testing with elderly people, we're testing with different ethnic groups, we're testing with people in different languages. How do they react to it? How can we make it relevant for the most amount of people?

J: I imagine the results you see are pretty widely variant depending on, not just the groups, but their personal preferences. Which might fall into some general categories based on the group they're in, but still, there's probably a pretty wide variance.

Brian: Yep. That two-way conversation and that dialogue is what we're doing a lot of work in right now. As I was mentioning earlier, with voice, with Watson technologies. Then also being, to your point, very, very careful about privacy. What do we do? How do we handle, if I create an ad for GSK, or a utility experience for Watson, that talks about cold flu, how do we handle information where people may be giving you medical information, be it voice or text, on an advertising experience? We have to follow a lot of rules and guidelines in privacy. There's also the GDPR, the Global Domestic Privacy Rules, are coming out, and they're launching out in May of '18. We're well ahead of that as well in already instituting all of those standards in our products, services, and experiences globally, so that we're able to really protect people's information and data, and they can trust that they're interacting with an experience that's not going to turn around and start selling personal information.

J: Right. You don't want to end up being a target, where they inadvertently told someone they were pregnant when they didn't even know yet.

Brian: Yeah. There's a lot of watch-outs and gotchas, because again, it's just such a brave new world that we're in. We've got to be very, very careful about how we're interacting with people and what we're doing with the information that they're giving us. Because it is now more than ever, as I was talking about, that cognitive system that's a two-way value exchange and communication. Where we need input and we get relevant output, and we have to be careful about what we do with the input and the output of that data. These experiences that we're building for brands, we need to be careful about the intellectual property. We build a cognitive experience. We're very protective of that intellectual property that that experience was built for that brand, but that doesn't then become part of another brand experience.

J: Right. I think that's probably a good lesson to take away, is that as you're designing any experience you have to think about both sides of the transaction. It's not just about what can you tell somebody, but what are you listening to, and how are you creating a dialogue, and making sure that what you're doing is actually adding value to their lives instead of just sticking something in front of their face.

Brian: Yeah.

J: Cool. Thanks for coming on the show today and talking a little bit about what you're working on. If somebody wants to learn more, or just reach out to you and get in touch, what's the best way to do that?

Brian: Sure, they can get me at Twitter, @bhull1, that's my handle on Twitter, and it'll be great to converse with anybody out there.