hocus pocus disappearing act with smoke
PHOTO: Aziz Acharki

Digital experience technology is at its best when it disappears. When it melts into the customer’s day. Its history has been one of incrementally removing friction from experiences until you forget the digital interface is there. Consider the humble checkout: from a multi-field form-fill for every purchase, to auto-populating upon login, to one-click, to a passing comment at a virtual assistant.

The e-shopping example is an embodiment of the pursuit of the digital experience holy trinity: convenience, speed and usability. We went from having to navigate a site and type out a bunch of information to not even having to take our phone out of our pockets. The technology interface fades a bit more with every seam removed. The furtherance of this disappearing act may be among AI’s greatest contributions thus far to the customer's ecommerce experience. And the customer, fittingly, is likely not aware of it.

As marketers we are probably too aware of AI by this point of perhaps peak hype. But marketers are buying, if not swallowing, the hype whole. We’re lured by dreams of automating boring tasks, drawing sharp insight from any random conglomeration of data, and finally delivering those true 1:1 digital experiences at scale. 

But how does the customer feel about it?

Mistrust of AI in the abstract abounds in the general public — from concerns over job security to high-profile technorati decrying AI as an Orwellian threat to humanity. In our day-to-day digital lives, however, AI mostly blends in when well deployed, and annoys us superficially when not (c’mon Alexa, when I ask to play "Hypnotize" I don’t expect Fleetwood Mac). Chatbots capable of natural language processing are the most tangible manifestation in AI in the eyes (and ears) of the customer, who doesn't have a nuanced understanding of the technology behind their experiences with brands. But other manifestations surface frequently, whether the customer is aware or not. 

Below are three primary ways customers benefit from marketing-related AI:

AI and: Customer-Centric Personalization 

Yes, personalized ads work for marketers’ conversion rates — but customers are just as likely to find these creepy and annoying as helpful. Personalization, like most all of marketing, works when it offers customers value. This may be anticipating and proactively delivering to a need, and thus save the customer time identifying, articulating and searching for it. For example, finding something to watch on a video streaming service can be frustrating. Content generally isn’t intuitively organized, search is slow and clunky (on television anyway, where most such content is consumed), and categorization can seem random. 

British telecom Sky’s video streaming service solves for connecting customers with content with an AI-driven recommendation capability. It personalizes not just for you or even your current mood but purports to recommend content that resides in the middle of the Venn diagram of whatever amalgamation of feelings you have at the time. Anyone who has given up searching and browsing a digital streaming service and retreated to the comfort of flipping channels on cable may understand why Sky developed a program to help you bypass its interface more quickly.

It is this type of personalization — customer-centric, not customer stalking — that has the potential to prove a loyalty-driving win-win.

Related Article: 3 Ways AI and Automation Aid Personalization 

AI and: Fluid Mobility

Mobile web traffic first surpassed desktop traffic in late 2016. Though mobile web, compared to desktop, offers a markedly limited browsing experience, the technology blends into our daily lives better. And that freedom made the experience favorable. Untethered to your static machine, or even a bulky laptop, you could be online anywhere.

With long page load times, poor responsiveness, and difficulty inputting commands, the mobile web experience, however, generally disappoints users. Mobile apps were meant to solve mobile experience challenges. Functionally different from websites, they’re purpose-built for specific conversions a brand needs to deliver efficiently. Good in theory, but in practice customers aren't inclined to download, maintain and remember to access hundreds of different little discrete interfaces on their phone. Research has shown under a third of apps will be used more than a handful of times, and customers will spend the vast majority of their time on only three apps. The rest will wait to be uninstalled. 

The answer to the app conundrum may lie in interface consolidation, as we’ve witnessed in China. The China-based super-app WeChat is a good example. It combines social media with messaging with payments with ride/food ordering with, um, the ability to “bottle” messages and throw them in a virtual sea, all under one mobile roof. Safe to say with one billion active users, this model has caught on in Asia. 

But even without superapps, AI has the potential to consolidate and dissolve mobile interfaces. Siri can get you an Uber without you even unlocking your phone. And of course, WeChat’s new Xiaowei virtual assistant will now let you bypass the WeChat interface and access all the services and content within.

Related Article: 3 Misconceptions About AI and Marketing

AI and: Humanity at Scale

In many industries, people still represent your brand to customers. Store associates, bank tellers, call center agents — an interaction with a human is worth its weight in web sessions in your customer’s estimation of your brand. But people are an expensive resource. In mid to low margin businesses, like much of B2C commerce, it’s not realistic to hire people to personally hold customers’ hands and guide them around their journey with your brand. So we let customers hunt and seek through various channels on their own.

And customers even like to navigate independently, as long as it’s convenient, fast and easy (yes, the digital experience holy trinity returns). But the brand gives up a significant degree of quality control over the experience, since many self-service channels customers make their way to aren’t controlled by the brand (review sites and resellers, for example). 

Here’s where AI bots show serious potential, if mixed results. The hype cycle started somewhere in mid-20th century science fiction, but when initially put into practice AI bots widely disappointed users. However, many of these colicky bot infants were the result of brands not understanding how to best develop and use chatbots. Now certain bright bot toddlers are emerging from this as best practices have developed and machine learning algorithms have been refined and duly trained. “Think [human] augmentation, instead of replacement,” Forrester says, while also noting that the brands who manage to get chatbots right achieve real business results.

And augmentation of humanity — not replacement for the sake of pure cost takeout — best serves the customer as well. Gartner predicted that by 2020, 85% of customer interactions could be handled by chatbots. We’ve already seen simple checkouts and inquiries duly handled by bots, but the experience is challenged when the task is complex. With human workloads free from routine customer interactions, however, bots can escalate complex interactions to humans. Humans in turn then (in theory) have the bandwidth to not just satisfy but exceed expectations for customers wanting or requiring special consideration. 

Moreover the bot itself is becoming more human and less another digital interface to get past. Now we are increasingly able to chat with bots via channels like voice assistant and messaging apps that don’t make customers go out of their way, as opposed to channels like browser chat windows and telephonic interactive voice response systems.

Customers may not be embracing AI at as rapid a clip as brands, nor should they. The customer’s sphere of concern regarding a brand centers on their overall experience with it, not the technology used to deliver that experience. However, to the extent that AI delivers value for the customer, and is not leveraged for customer-oblivious (or averse) reasons such as cost takeout or the urge to chase a shiny new thing — they won’t think about your AI enough to have a strong opinion as they did with the rejection of early chatbots. The AI will simply blend into their daily lives, taking your brand with it.