How NVIDIA Uses AI to Boost Digital Marketing Efficiency

Adobe Experience Platform

By Mon­i­ca Lay, Group Prod­uct Mar­ket­ing Man­ag­er, Adobe Expe­ri­ence Plat­form  

NVIDIA’s graph­ics pro­cess­ing units (GPUs) deliv­er the com­put­ing capa­bil­i­ty nec­es­sary to pow­er every­thing from graph­ics-inten­sive games to self-dri­ving cars and AI devel­op­ment work. In recent years, this leader in AI has looked to AI-dri­ven tools to serve its own cus­tomers as well. 

This includes NVIDIA’s mar­ket­ing organ­i­sa­tion, which turned to Adobe’s Cus­tomer Expe­ri­ence Man­age­ment (CXM) solu­tions that use AI to assess the effec­tive­ness of cam­paigns and bet­ter antic­i­pate cus­tomer behav­iours. Dri­ven by data analy­sis, the approach sup­ports busi­ness resilien­cy through enhanc­ing cus­tomer expe­ri­ences, dri­ving greater effi­cien­cies online and align­ing the impact of mar­ket­ing on sales.  

When Adobe launched a beta pro­gram for AI-dri­ven Intel­li­gent Ser­vices to address these kinds of needs, NVIDIA’s data-dri­ven team was one of the first to sign up. In fact, we’ve been work­ing togeth­er on AI for a while, and sep­a­rate­ly have more to share on our work togeth­er today with NVIDIA’s Spark 3.0 news. Here, we’re shar­ing some ear­ly results of how AI is help­ing NVIDIA enhance the cus­tomer experience.  

AI-Dri­ven Mar­ket­ing Attribution   

Most mar­ket­ing teams have one under­ly­ing chal­lenge in com­mon: prov­ing the effec­tive­ness of mar­ket­ing chan­nels and cam­paigns. The issue has been exac­er­bat­ed in the last cou­ple of years, as brands began to invest in every­thing from email mar­ket­ing and web­sites, to social media and emerg­ing chan­nels like voice-enabled AI.   

In a recent exam­ple, NVIDIA want­ed to under­stand how mar­ket­ing impact­ed the sales of a new line of GPUs. With Attri­bu­tion AI (part of Adobe’s Intel­li­gent Ser­vices), the sys­tem ran an in-depth analy­sis across all rel­e­vant data sets and iden­ti­fied chan­nels (paid media vs email) and cam­paigns (bun­dled offers vs prod­uct refresh) that proved to be most effec­tive. Incred­i­bly, it helped them tie mar­ket­ing clos­er to sales by point­ing out incre­men­tal lift and quan­ti­fy­ing the sales (or con­ver­sion) that would be missed with­out marketing. 

Attri­bu­tion AI also enabled NVIDIA to move beyond exist­ing forms of attri­bu­tion, where often­times too much cred­it is giv­en to the “last-touch,” – the final engage­ment point before an actu­al pur­chase or con­ver­sion event hap­pens. In tra­di­tion­al rules-based mod­els, dif­fer­ent touch points in the cus­tomer jour­ney are often mis­rep­re­sent­ed and under­val­ued or over­val­ued as a result.   

Recent­ly, the mar­ket­ing team was tasked with dri­ving event reg­is­tra­tions. They want­ed to ensure that funds were being spent wise­ly and had the biggest impact. With Attri­bu­tion AI, they were able to see the effec­tive­ness of invest­ments across chan­nels, while drilling down into spe­cif­ic cam­paigns. In this case, they opti­mised paid search around a set of key­words – an effort that drove a 5X lift in paid registrations.   

Pre­dict­ing Cus­tomer Behav­iours with AI     

In today’s com­pet­i­tive busi­ness envi­ron­ment, reduc­ing churn has become the new growth dri­ver for many com­pa­nies. Brands need an abil­i­ty to antic­i­pate cus­tomers that may jump ship and look for ways to re-engage and retain them. NVIDIA lever­aged Cus­tomer AI in its con­sumer busi­ness to bet­ter iden­ti­fy those who were like­ly to stop using the ser­vice. With a remark­able 96% accu­ra­cy rate, the ser­vice analysed every­thing from the num­ber of days spent between ses­sions to usage fre­quen­cy. AI was able to pro­vide a seg­ment of cus­tomers who would be best served through spe­cial­ly tai­lored con­tent and campaigns.   

Accord­ing to Derek Sun, dig­i­tal mar­ket­ing ana­lyst at NVIDIA: “AI and machine learn­ing have a big role to play in the future of mar­ket­ing, as busi­ness­es want to bet­ter under­stand their cus­tomers and make oper­a­tions more effi­cient. Adobe’s Intel­li­gent Ser­vices pro­vide mature data sci­ence mod­els that are both flex­i­ble and fast to deploy, help­ing us be more in tune with our com­mu­ni­ties and deliv­er the right con­tent to the right per­son at the right time, while improv­ing the effi­cien­cy of our mar­ket­ing investments.”   

NVIDIA is a great exam­ple for brands and mar­keters in any indus­try that are explor­ing ways in which AI/ML can help their organ­i­sa­tion and dri­ve lead­er­ship in CXM. Across all the com­mu­ni­ties NVIDIA serves, from gamers, artists/designers, researchers, data sci­en­tists, and inno­va­tors across all indus­tries, the com­pa­ny is show­ing how data and AI can help ensure that every deci­sion around cus­tomer engage­ment and mar­ket­ing invest­ment is the most impact­ful, most cost-effec­tive choice possible. 

To learn more about Adobe’s Intel­li­gent Ser­vices and how they can help your organ­i­sa­tion see quick­er val­ue from AI, vis­it here


Adobe Experience Platform
Digital Europe

Posted on 05-14-2020


Comments

  • By Hodor Hanner - 2:53 PM on June 18, 2020   Reply

    Attri­bu­tion AI also enabled NVIDIA to move beyond exist­ing forms of attri­bu­tion but not only Nvidia. All the tech­nol­o­gy busi­ness this days try to switch on AI for bet­ter performance.

Join the discussion