Back to Gartner for Marketers Blog

Hype Detected: Conversational Marketing and Artificial Intelligence

By Benjamin Bloom | August 01, 2018 | 0 Comments

MarketingTechnology and Emerging Trends

In a memorable exchange from Terminator 2: Judgement Day, young John Connor (dirt-bike rider and hacker extraordinaire) tells his Terminator body guard about the things a Terminator must do to seem more human. In a deleted scene (brought back in an extended cut) John and Sarah perform minor surgery to remove a governance setting and disable “Read Only” mode, to allow the Terminator 800 to “not be such a dork all the time.” T800_Port_Open


With the modification made, the Terminator then  learns the slang that will help him “blend in” – including the memorable, “Hasta La Vista, Baby.”

In 2003, the American Film Institute ranked The Terminator from the 1984 original as #22 on its list of villains, and the reprise in T2 as number 48 on its list of Heroes.    In my early experiences with AI, it was hard to see either:  engaging with natural language personas like Eliza or the Instant Messenger bot Smarterchild over my 14.4kbps modem was mostly a harmless game. By contrast, our interactions with other humans, even in the chat window of Warcraft II, were real and interesting.

Modern consumer messaging services such as Facebook’s Messenger, Apple’s iMessage, WhatsApp (Facebook), and Skype (Microsoft), are a new surface for AI to operate on, and we see a tension driven by the transition from consumer use to commercial. While consumers now see businesses in their home turf, marketers hope to use novelty, convenience and effectiveness to bring conversational engagements into the mainstream.  

Support for a mix of human and artificial intelligence (AI)/bot conversations is central to the value proposition of conversational marketing, which makes its first appearance on the Hype Cycle for Digital Marketing and Advertising 2018.  Through persistent, session-based,cross-channel exchanges  – across messaging platforms, native applications or smart devices-  brands engage with customers in natural language, using a blend of text and audio.  Artificial intelligence is a key component that keeps these interactions informal and…conversational. Conversational marketing is in the Innovation Trigger stage approaching the peak of inflated expectations.

Consumers are still processing, even as they cannot escape media portrayals of Alexa, Google Assistant, or Siri pushing their associated hardware ecosystems.  Despite explosive growth fueled by consumers buying these devices for every room in the house, a recent study found that 56% of respondent believe that increased regulation of AI is a necessary measure.

Some AI conversations take the form of adorable pilot programs/PR moves like Connie, the  Robot Concierge at Hilton McLean Virginia which leverages IBM’s Watson.  Or in a confluence of media, comedy and technology, Microsoft’s social AI “Zo” interviews the host of a comedy podcast (episode sponsored by Microsoft, some language on this podcast is NSFW).

In response, growing evidence suggests that retailers are looking to personalization instead of replacing humans with robots. Yet conversational marketing often makes it hard to tell if you’re engaging with a robot – and many times you are not. The clienteling applications of conversational marketing, in which rich cross platform messaging facilitates “concierge commerce” lifted conversions at one retailer by 54%.

Either way, consider (hopefully in a heroic application) the idea of Cybernetic CX. Cybernetic naturally refers to a combination of human and machine, like the skin and muscle over the Terminator’s metal endoskeleton. Through this approach, marketers can think of applying AI in four steps: Anomaly Detection, Problem Diagnosis, Determine the Remedy, and Application of the remedy.

Detection: complex computational resources can ingest vast amounts of normalized data, and immediately surface new segments worth targeting.

Diagnosis: likely needs are parsed to identify  a cross-sell opportunity, likelihood to churn, or an advocate to be rewarded.

Determination: treatment option(s) are selected, or a test constructed to discover one from thousands of potentials.

Application (of the remedy): campaigns or messages are launched as needed, to narrow segments with individualized offers.

This process may be happening in real time as you converse with your friends Alexa, Siri, or Cortana…or as your favorite retailers update their Lifetime value models in advance of Black Friday, taking into account what you bought last year, last month, and last week, to inform their merchandising strategy and develop new “doorbusters.”  As algorithmic marketing tests new concepts, you may even find the items of your “wish list” neatly boxed and lying at your doorstep, with an offer to be charged only for what you keep.

In consumer marketing, our choices are rarely so grave as the battle to save the hero of The Resistance, but business imperatives and hype can still make us feel we are being chased by an unrelenting trend.  Are the conversations handled by our algorithmic aldermen to be valued as highly as those with our parents or peers? It will most certainly be up to us, as Andrew Frank’s 2018 DMC presentation, “What AI Teaches Us About Being Human” lays out:  we must “design organizations that augment, rather than replace human capabilities and keep humans in the loop.”

As foreign or frightening as AI can be, in text, video, or in the flesh, culturally resonant moments will stem from a mix of scientific progress and human emotion.  Did you know that “Hasta la vista, baby” came in at #76 on the AFI’s 100 Years…100 Movie Quotes? Perhaps in our conversations with machines in web, voice, social messaging apps, SMS, we will similarly find heroes who come to our aid and most importantly, speak our language.

Leave a Comment