AI in Marketing: Long Live the Customer Experience Revolution

The word “rev­o­lu­tion­ary” might be one of the most overused adjec­tives in mar­ket­ing. What does “wid­get x” or “ser­vice y” have in com­mon with the bloody, world-chang­ing events that took place in coun­tries like France or Tsarist Rus­sia? That said, when it comes to apply­ing Arti­fi­cial Intel­li­gence (AI) in mar­ket­ing, I think the adjec­tive may, for once, be appro­pri­ate. AI, like the rev­o­lu­tions before it, will bring about a com­plete change of ideas, lead­er­ship and direc­tion. It will trans­form every­thing we know about mar­ket­ing (although hope­ful­ly it will accom­plish this with­out guil­lotines and gunfire!).

So, here we go, I’m going to say it: AI will rev­o­lu­tionise marketing.

Demys­ti­fy­ing AI

On a tech­ni­cal lev­el AI isn’t the eas­i­est thing to get your head around—unless you’re a com­put­er sci­en­tist. For mar­keters, how­ev­er, the vital issues are Machine Learn­ing (ML) and AI appli­ca­tions. ML refers to algo­rithms that are pro­grammed to learn from data, where­as AI refers to appli­ca­tions that mim­ic human cog­ni­tive abil­i­ties, for exam­ple apps that can recog­nise images or respond to cus­tomer queries.

The rev­o­lu­tion­ary bit comes when these two tech­nolo­gies com­bine. ML is like the great rev­o­lu­tion­ary thinker: it comes up with insights and real-time knowl­edge, and ways for the AI sys­tem to keep improv­ing and grow­ing in sophis­ti­ca­tion. ML feeds into the AI application—the agent of change—whether that’s a chat­bot vir­tu­al assis­tant or a pro­gram­mat­ic algorithm.

AI in marketing

What sorts of changes are we talk­ing about? One of the most excit­ing ways that AI will trans­form mar­ket­ing is its abil­i­ty to cre­ate hyper-per­son­al­i­sa­tion. Per­son­al­i­sa­tion holds the key to cus­tomer loy­al­ty, enabling brands to build real, long-last­ing rela­tion­ships with cus­tomers as indi­vid­u­als. Until AI how­ev­er, true per­son­al­i­sa­tion hasn’t real­ly been pos­si­ble, because per­son­al­i­sa­tion at scale has been impos­si­ble to achieve with­out rack­ing up the sort of costs that make even the hardi­est CFO burst into tears. Per­son­al­i­sa­tion requires sift­ing through vast amounts of data in real-time—a func­tion that the human brain can­not perform.

This is where AI tech­nol­o­gy proves its worth. As it’s based on high-pow­ered com­pute capa­bil­i­ties, AI is a lot smarter than us at cer­tain tasks. Com­pa­nies are build­ing AI sys­tems that can mim­ic human con­ver­sa­tion and “work” with cus­tomers to resolve issues, triage com­plaints and queries, and pro­vide high­ly per­son­alised recommendations.

But it’s not all about replac­ing us humans. AI will help us do our jobs bet­ter. One great exam­ple of this is Cog­i­to. Cogito’s tech­nol­o­gy per­forms in-call voice analy­sis, pro­vid­ing real-time behav­iour­al insights to help com­pa­nies’ agents bet­ter engage with cus­tomers. It’s pret­ty cool to be able to read how your cus­tomers are feel­ing to help ensure you’re giv­ing them the best ser­vice pos­si­ble. It’s the sort of change that deserves to be called revolutionary.

The cur­rent state of play

Where do we stand with AI today? There are two sides to this ques­tion: what’s pos­si­ble, and what peo­ple are actu­al­ly doing. On the pos­si­ble side, there has been a huge surge in AI appli­ca­tions. Here at Adobe, for exam­ple, we’ve pio­neered an AI expe­ri­ence sys­tem, Adobe Sen­sei, which uses AI and ML to help busi­ness­es dis­cov­er new insights that would oth­er­wise have been hidden.

The sys­tem draws on mas­sive vol­umes of con­tent and data assets to gain deep cus­tomer insights and deliv­er high­ly per­son­alised expe­ri­ences for cus­tomers. HSBC, for exam­ple, has test­ed Sensei’s abil­i­ty to per­son­alise the order of prod­ucts shown on a web­site to draw in traf­fic based on indi­vid­ual cus­tomer pref­er­ences. In its tri­al, HSBC used the tech­nol­o­gy to boost a prod­uct that per­formed well on its web­site, lead­ing to a 109 per­cent increase in cus­tomers reach­ing that product.

Else­where in the world of mar­ket­ing, AI is hav­ing one of its great­est impacts in the rise of voice-enabled per­son­al assis­tants. These “chat­bots” demon­strate the pow­er of AI to con­nect cus­tomers with auto­mat­ed per­sonas that use AI-derived insights to make per­son­alised rec­om­men­da­tions. Take Pefin’s new AI finan­cial advi­sor, which was pre­viewed at SXSW 2018. This chat­bot is so sophis­ti­cat­ed it can help users nav­i­gate life’s key finan­cial decisions—everything from get­ting a mort­gage to sav­ing for retire­ment. This is the sort of high­ly-per­son­alised expe­ri­ence that once would only have been pos­si­ble through a (very expen­sive) human agent, and high­lights the pow­er of AI to pro­vide bespoke expe­ri­ences at scale.

That’s what’s pos­si­ble. Let’s turn our atten­tion to what’s actu­al­ly hap­pen­ing. To what extent are busi­ness­es embrac­ing AI appli­ca­tions? Well, the answer depends on who you ask. Accord­ing to For­rester, just 11 per­cent of brands can be con­sid­ered AI experts, while more than half fall into the “novice” or “lag­gard” cat­e­go­ry. But is that real­ly sur­pris­ing? The tech­nol­o­gy is still very new, and with such a big change most busi­ness­es would be wise to con­sid­er exact­ly how to deploy the tech­nol­o­gy to get the most value.

Per­haps a more reveal­ing stat is this: 84 per­cent of mar­ket­ing organ­i­sa­tions are imple­ment­ing or expand­ing AI and machine learn­ing in 2018. Even if most busi­ness­es cur­rent­ly lag in AI adop­tion, you can bet your bot­tom dol­lar that won’t remain the case for long.

What does this mean?

Some of you may remem­ber the debate a few years ago about whether cus­tomer loy­al­ty had had its day. The argu­ment ran that mod­ern cus­tomers are too empow­ered and have too much choice. We’re fick­le and won’t hes­i­tate to switch brands. For me, the true pow­er of AI in mar­ket­ing is that it revers­es this trend, mak­ing cus­tomer loy­al­ty a real­is­tic goal once more. AI allows brands to know their cus­tomers as peo­ple, and inter­act with them with the sort of insight you would expect from a best friend. If that doesn’t build loy­al­ty, I don’t know what will.

We’re at a crit­i­cal junc­ture with the deploy­ment of AI. The use cas­es are firm­ing up and the ben­e­fits of the tech­nol­o­gy are clear to see, but many firms have yet to make a play. My advice is sim­ple: act now and don’t be left behind. Busi­ness­es that mas­ter AI-enabled mar­ket­ing first will be at a sig­nif­i­cant com­pet­i­tive advan­tage. Vive la rev­o­lu­tion! Let’s get busy rebuild­ing marketing.