Photo: IBM
Will shoppers want to interact with AI ads?
The Weather Company, acquired by IBM in January, will begin rolling out digital ads that people can communicate with this fall. Powered by IBM Watson’s artificial intelligence, consumers will be able to ask questions via voice or text and receive relevant information about the product or offering.
The first companies to tap Watson ads will be Campbell Soup Company, GSK Consumer Healthcare and Unilever.
Unlike ads tied to keyword searches, Watson Ads are able to pick up the nuances of natural language.
For example, a consumer seeing a Campbell’s digital ad may ask by voice interaction, “What can I make for dinner tonight?” Using machine learning and the data ingested, Watson can sort through ingredient and flavor profiles to make recommendations based on the weather, time of day, location and even ingredients users have on hand.
In another example, an allergy medication brand may address questions about whether the product is appropriate for children or its side effects.
In its responses to queries, Watson Ads will also tap Weather’s Co.’s reach into historical weather data, current conditions like temperature, precipitation or the current weather forecast to help individuals “make decisions quickly and directly in the ad unit without leaving the experience.”
However, the technology will be fighting consumers’ general apathy toward digital ads, especially on mobile, although the popularity of Apple’s Siri and Amazon’s Alexa suggests an opening for artificial intelligence.
“It’s going to be the way we interact with multiple devices in an Internet of Things world,” Carl Fremont, global chief digital officer at MEC, Unilever’s media agencies of record, told The Wall Street Journal.
Discussion Questions
DISCUSSION QUESTIONS: Will consumers want to interact with digital ads enabled by artificial intelligence? Will the AI engagement help overcome consumer aversion to digital ads?
Poll
BrainTrust
Chris Petersen, PhD.
President, Integrated Marketing Solutions
Brian Numainville
Principal, The Feedback Group
The challenge may well be that mobile devices are already being clogged with digital ads. The AI won’t matter if the customer doesn’t notice the ad. One could argue that AI in retailer apps would be better utilized, but the research shows that customers only use about five apps on a regular basis. The key for AI would seem to be the portals for search that customers use most.
IBM is at the sweet spot of harvesting all that the company promised 20 years ago when their computer won a chess game. Their integration with the Weather Channel will unlock even more possibilities than digital ads – much more. Exciting to see how Watson will expand and learn.
If digital ads can provide relevant information quickly and concisely, they could successfully become part of a brand’s advertising tool kit. My questions is: once a consumer has the information she requested, will she remember which brand delivered the information?
I’m not sure AI will add enough to make shoppers any more interested than they already are or are not. Everything we’ve seen so far says interest in ads may be a personality trait (or flaw), not a technology-driven issue.
I believe that voice-activated technologies will be the next wave of consumer communication. I signed up to Beta Amazon’s Echo — incorrectly assuming that a. the interface wouldn’t work and/or; b. the novelty would wear off fast.
The truth is that the notion of “hands-free” communication actually is a solid one once you get used to it. The key to mass adoption will be functionality and interface. The more natural the interface, i.e., how close it comes to mirroring human speech, the wider the acceptance. And obviously the more value the interface creates, i.e., Watson’s more nuanced answers versus Echo’s limited functionality, the more often people will be inclined toward signing on for interactive communication.
But, as in all things media, in the end content is king. If it is useful it will grow. If not, it will be just another short-lived gimmick.
I watch my kids, 11 and 14, and I think that generation will be much more voice, video and hands-free oriented than any of the rest of us. They barely learned how to write before being given laptops for school. The only papers they turn in are for math. They are truly the YouTube and Siri generation. In fact, they get annoyed with me when I don’t use the voice features of our car — which they, I confess, know how to use better than I do.
These new capabilities should help make ads far more relevant and engaging than in the past. Consumers have way too short of attention spans, especially with digital marketing, to care about anything that is not directly applicable to them. These messages also have to be extremely intuitive and allow for seamless engagement. If it takes too much effort, all with be lost. This new offering to the marketplace shows great signs of adoption so far, and I see AI augmented by machine learning becoming a more and more pervasive way to reach shoppers in the most personal ways.
To the extent that customers are willing to interact with ads, I think this will make ads more engaging and perhaps more valuable or useful for consumers. The problem is this doesn’t change the fundamental problem of ads, which is that they are an interruption play in a world where consumers increasingly look to avoid such interruptions.
What I would rather see IBM do with The Weather Channel assets is actually predict the weather better. Or at least let me interact with the app in a more natural-language way: “I want to go for a hike in an hour. What are the odds I’ll get rained on?” THAT would be more interesting to me. If Watson can build enough trust with consumers around the weather, then maybe consumers would be willing to take recommendations too — “It’s going to snow today. How about some soup for dinner?”
But making consumers wade through a bunch of ads they haven’t asked for — however smart and chatty those ads might be — that’s a no-go for me. Sure, some people will love them. But it’s still an analog way of advertising, no matter how much AI is behind it.
Being clogged with ads does not bode well for ads in a different format. If the ads themselves are clever, get noticed by consumers, let consumers know they can ask questions and are able to get their questions answered, then this format could be successful. There are a lot “ifs,” however.
Let the Millennials play with Watson. I donate my $10 a year to WeatherUnderground.com — part of The Weather Company and have NO ADS to bug me.
AI will have a bump up and then fade. Soon it will not be hip talking to your TV remote or your phone for help. Are we really so stupid we need Watson to help us with ideas? Let the computers calculate and process. I can ask them questions via a keyboard. They are not and never will be human … just more tech trends for super trendy people. All I need is my CLAP ON CLAP OFF light switch.
Google has been doing this for years. It’s the reason I own a Droid. My phone can literally be in my purse and I can say “OK Google, find me a coupon for Macy’s,” and it will. When I lose my phone I say “OK Google, give me the definition of liberty.” By the time it’s done giving me the definition, I have it in hand.
However, I’m not sure how that will work with ads that weren’t requested. They tend to be annoying. I view ads much like advice. Unsolicited advice is rarely welcome. Neither are unsolicited ads in most cases. I think it will be all about timing and circumstance. If I walk into my local drug store and a coupon for Tums pops up, I might be interested. That same ad if I’m watching a game or at a graduation party will annoy me.
Watson is an extremely cool technology. I’m looking forward to seeing what more IBM will do with it!
The key to mobile advertising is compelling and relevant content and not too many ads. Mobile screen real estate is easy to clutter. If the AI can help deliver real-time, relevant ads there is a better chance of conversion and conversation. I can’t tell from the article whether Watson AI is determining what ad to deliver or handling the interaction or both.
I’m quite skeptical. I have spent years researching consumer interests around cooking and food. The example shown here for interacting with the ad is quite frivolous.
It’s suggestive of a classic tech problem: Engineers, who have built a clever technological gizmo, desperately searching for a reason for their technology to exist.
Simply can’t imagine that AI “engagement” will shift the general resentment of digital ads. As I wrote elsewhere over the weekend, personalization (engagement) is in the eye of the beholder. Unfortunately, the “engagement” that drives profit for advertisers is NOT the engagement consumers want.
As marketers, words like “engagement” and “personalization” are being massively abused. When we sit back and consider what they mean to a real human being, that’s very different from what Watson could ever offer (and I’ve been tracking AI since the 1980s).
The first challenge will be breaking through all of the ad clutter to get people to notice the AI ads. And if that happens, the further challenge will be to keep people interacting with AI ads once the novelty wears off if they found them engaging enough. Hard to tell where this one might go.
Why do we always think sell, sell, sell, whenever cool new technology comes up. Haven’t we learned yet that folks, especially Millennials, don’t want to be sold to? It’s serve before you have permission to sell. Nikki’s thoughts about using the Weather Channel along with Watson are spot on.
And that’s my 2 cents.
The psychology of this is fascinating. Watson enables us to think about these interactions not like an ad, but a conversation. The old metaphor of brand and consumer being in dialog is becoming a very real thing.
These early experiments will likely flesh out who we trust; the product (Campbell’s Soup) or the aggregator (Kroger, Alexa, Google Now). When our intrepid customer is shopping — in a store or on her phone at lunchtime — does she feel she has a trusted companion with her, or a salesperson hounding her?
I would expect Watson is working on the very question we are trying to answer today. Historic data and that from focused experiments with different types of machine learning will yield the answer.
This just might be the next generation of self-check out.
There’s no doubt it’s a very cool technology. However, once the novelty wears off, I’m skeptical that the engagement factor will be enough to overcome consumer aversion to ads.