IBM have announced it will use cognitive computing platform Watson to “speak” with consumers through voice-activated advertising units. But do customers really want a conversation?
Odds are – substantially – you don’t click on ads. For some, odds are they don’t even see the ads, or are taking steps never to see ads. Is the issue that the ads just don’t talk to you? IBM Watson and their new Weather Co acquisition are betting you just need to talk (to the ads).
Voice interaction with display media isn’t new. Since HTML5 we’ve been able to initiate a microphone connection on third party sites. But it never really took off. Why now?
Adding voice commands to ads for Campbell Soup, Uniliver and GlaxoSmithKline, powered by smart AI (and Watson is amongst the best), is an exciting media product innovation for Weather Co.
What the change means
A.I is hot right now and it’s certainly a good a stunt. Wisely timed to demonstrate value behind Weather Co’s purchase by IBM. CEO Ginni Rometty is saying Watson is their growth focus – and some say she needs to be right.
AI promises more human-like interaction with technology. This year has seen the rise of conversational interfaces such as branded chatbots like KLM’s FB Messenger, Poncho’s weather assistant and of course Siri, Alex, Cortana and Google Assistant. 3.1 billion internet users spend more time chatting in messaging apps than any other activity online or on mobile.
Implications for marketers
The experiment here is whether voice will solve the low interaction rates with online ads and ads will solve low interaction rates with voice. Digital media interaction rates on third-party sites are dismal – under 0.5% in N.America (Sizmek, 2015). And Siri (the widest used voice-command technology) has never been used by 85% of iPhone owners. Apple has mentioned Siri serves 1 billion requests a week. There are about as many iPhone owners, so on average, once each per week. By comparison, iPhone users open apps approximately 80 times a day.
The Watson-powered ads promise users will be able to ask them more questions, such as a CPG product’s suitability for allergies or children. Currently, marketers need people to click on an ad, and then click around, to find that out. Current behaviour says they’ll almost never ever do that. Current behaviour says they’d prefer to use chat. Will results of Watson’s and Weather Co’s experiment reveal they just wanted to talk?