Personalisation Technology: Automated Personalisation

So, your busi­ness has adopt­ed per­son­al­i­sa­tion, and you’re on the path to cre­at­ing new bonds with your con­sumers by under­stand­ing their pref­er­ences and desires on the macro scale. That’s excel­lent- even a small invest­ment in the world of per­son­al­i­sa­tion can pay mas­sive div­i­dends, widen fun­nels and lock down seg­ments. So what’s next?

There tru­ly is no end to the poten­tial inno­va­tions of an ambi­tious, grow­ing per­son­al­i­sa­tion regime, but there is a lim­it to our own endurance. It’s easy to be over­whelmed by the sheer pos­si­bil­i­ties, the breadth of the data pro­vid­ed by this analy­sis-rich approach to mar­ket­ing and con­sumer rela­tion­ships. That’s where auto­mat­ed per­son­al­i­sa­tion comes in.

Auto­mat­ed per­son­al­i­sa­tion builds out of the frame­work we’ve estab­lished in our pre­vi­ous attempts to reach our cus­tomers more per­son­al­ly. It takes on forms you should be quite com­fort­able with by now- still tweaks the con­tent and appear­ance of our web and mobile pres­ences to reflect con­sumer desires and pref­er­ences, but most crit­i­cal­ly, it then applies the insight we might glean from these inter­ac­tions and improves itself auto­mat­i­cal­ly. There will always be a place for human insight in mar­ket­ing, of course, but the ben­e­fit of automa­tion is that it’s both con­stant and imme­di­ate. There is sim­ply no alter­na­tive for rapid reac­tion to shift­ing mar­ket trends- auto­mat­ed per­son­al­i­sa­tion has the capac­i­ty to respond to often imper­cep­ti­ble shifts in cus­tomer behav­iour with­out any prompt­ing. The poten­tial speaks for itself. Not only does it assist us in the heavy lift­ing of analy­sis, adjust­ment and deploy­ment, it also means that when we do stop to assess what we’ve learned from a cam­paign or a trend, we have not only the ini­tial data point but an entire process of adap­ta­tion that can serve to either con­firm or deny a per­spec­tive we’d oth­er­wise be just attempt­ing to test ourselves.

More to the point, auto­mat­ed per­son­al­i­sa­tion is the only oppor­tu­ni­ty for 1:1 mar­ket­ing at scale, a sys­tem that gen­uine­ly allows spe­cif­ic inter­ac­tions with sin­gu­lar cus­tomers at their point of entry, and all the way through the fun­nel beyond. That’s a pow­er­ful oppor­tu­ni­ty, but it’s also a demand­ing one. But what do we need as base­line from an auto­mat­ed per­son­al­i­sa­tion solution:

1. Trans­paren­cy: We need a firm hand on every­thing that is being done with our data and our seg­ments. That means total trans­paren­cy- not just what is being done, but _why _it’s being done. Are there insights that might not jump out from the raw data? Are there prod­uct affini­ties that might not be clear at first sight?

2. Ease of Imple­men­ta­tion: Per­son­al­i­sa­tion and auto­mat­ed per­son­al­i­sa­tion in par­tic­u­lar are com­plex enough with­out chal­lenges in imple­men­ta­tion. That’s why Adobe Tar­get includes visu­al cam­paign set­up and core ser­vices that make shar­ing cre­ative and test ideas extreme­ly easy.

3. Ease of Con­fig­u­ra­tion: Pre­vi­ous­ly, users could expect to use a rel­a­tive­ly sim­plis­tic resid­ual vari­ance mod­el for tar­get­ing, one still suit­able for straight­for­ward cam­paigns with few vari­ables, but it’s impor­tant to have the capac­i­ty to plug in oth­er mod­els. Adobe has been work­ing hard to pro­vide a ran­dom for­est mod­el, an advanced and high­ly accu­rate approach that lever­ages insight on oth­er con­sumers to help the machine deter­mine the best course of action in the present, and life­time val­ue model.

4. Ease of Exten­si­bil­i­ty: Of course, new algo­rithms are con­stant­ly being devel­oped. Adobe’s got some in the works right now. A flex­i­ble approach to auto­mat­ed per­son­al­i­sa­tion means being able to update the mov­ing parts when a bet­ter solu­tion comes around. Adobe Tar­get allows end users to plug in their own algo­rithms if nec­es­sary, as well as pro­vid­ing a pro­file API that con­nects with CRM and enter­prise data to deter­mine if it’s rel­e­vant and pre­dic­tive: Exten­si­ble automa­tion is robust automation.

5. Ease of Main­te­nance: It’s impor­tant that test­ing new com­bi­na­tions, adding new options, tweak­ing exist­ing cam­paigns and ensur­ing every­thing plays nice­ly togeth­er is easy. This is anoth­er exam­ple of reduc­ing the amount of work on your end- the major sell­ing point of automation.

A great exam­ple of a com­pa­ny using auto­mat­ed per­son­al­i­sa­tion through Adobe Tar­get is Sky UK. Their goal for the Sky Shop was to use data to deliv­er rel­e­vant con­tent and mes­sag­ing to users engag­ing with the shop in order to dri­ve incre­men­tal sales and upgrades on desk­top and mobile. In order to do this they worked with their part­ner DBi and the Adobe Mar­ket­ing Cloud to accu­rate­ly iden­ti­fy spe­cif­ic seg­ments of cus­tomers and deliv­er per­son­al­ized offers and mes­sages based on behav­ior and prod­uct hold­ing. This approach enabled them to deliv­er 2.7 mil­lion per­son­al­ized expe­ri­ences and sig­nif­i­cant incre­men­tal sales uplift. The Adobe Tar­get auto­mat­ed per­son­al­i­sa­tion report­ing also enabled them to uncov­er hid­den affini­ties between prod­ucts that were not evi­dent when man­u­al­ly tar­get­ing or ana­lyz­ing the data. You can hear more about their work here: https://www.youtube.com/watch?v=bQ5Il2-7wBw

Sky UK’s use of auto­mat­ed behav­iour­al tar­get­ing moves us clos­er to our vision of dri­ving real val­ue from our per­son­al­i­sa­tion tech­nol­o­gy invest­ments: cus­tomers get more rel­e­vant, engag­ing con­tent; the busi­ness gets more rev­enue from and clear return on invest­ment; and our peo­ple and process­es ben­e­fit from more time to think about the longer-term goals of our dig­i­tal expe­ri­ences, rather than being locked up in cre­at­ing man­u­al busi­ness rules or run­ning reports all day.