I have the honor of trav­el­ing the world to meet with opti­miza­tion teams and com­pa­nies across a myr­iad of industries—retail, travel, finan­cial ser­vices, auto­mo­tive, high tech, etc.—to dis­cuss dig­i­tal busi­ness strate­gies, dig­i­tal mar­ket­ing orga­ni­za­tions, and opti­miza­tion pro­gram devel­op­ment. Glam­orous, I know, but the real­ity of these speedy trips is that I spend a lot of time work­ing out of hotel rooms. The truly excit­ing part of these trips is see­ing the expo­nen­tial value of test­ing and tar­get­ing being real­ized more effi­ciently than ever before. This is the result of opti­miza­tion teams gain­ing access to the enhanced inte­gra­tions between pre­vi­ously siloed pro­file data sets, lever­ag­ing the ease of use of inter­faces, and using a breadth of both man­ual and auto­mated means for effi­ciently and accu­rately deter­min­ing the rule sets that will drive rel­e­vant expe­ri­ences across a diverse vis­i­tor population.

Travel and hos­pi­tal­ity cus­tomers can now “live the dream” of lever­ag­ing pro­gres­sive mul­ti­chan­nel pro­files that adapt based on whether cus­tomers are trav­el­ing for busi­ness or plea­sure. Cus­tomers receive offers and in-flight screen con­tent that incor­po­rate the most desir­able per­sonal crea­ture com­forts and acces­sories based on the multi-channel data aggre­gated and reg­u­larly updated in their pro­file. Finan­cial insti­tu­tions can import daily feeds of pro­files aggre­gat­ing branch, call cen­ter, and other offline data with online first-party, and even third-party, data feeds includ­ing valu­able vari­ables, such as credit rat­ing, that help score and con­nect the most appro­pri­ate offers to dis­play at first touch based on pre­qual­i­fi­ca­tion and level of investment.

And yet, I still hear on a semi-regular basis from cer­tain teams that these per­son­al­iza­tion sce­nar­ios seem out of reach, or “beyond what we’re able to achieve with our con­ser­v­a­tive resources.” Pro­grams are left mired in inci­den­tal or iso­lated test­ing of indi­vid­ual high-value ele­ments, such as the home­page hero ban­ner at a pop­u­la­tion level, with­out any analy­sis of how these tar­geted expe­ri­ences might influ­ence key sub­seg­ments within the pop­u­la­tion or affect expe­ri­ences fur­ther down the fun­nel. This not only severely lim­its the poten­tial returns that a team can achieve with its test­ing pro­gram, but also affects the behav­ior and loy­alty of sub­seg­ments that receive a dis­jointed or imper­sonal expe­ri­ence, opens your site and expe­ri­ences to the risk of fric­tion, frus­tra­tion, and churn, and it can hin­der your pro­gram growth and strat­egy because it’s based on improper analy­ses of the rel­a­tive pre­dictabil­ity of variables.

The excuse is often that there are only sin­gle, part-time, or few ded­i­cated resources avail­able to exe­cute tests. In our expe­ri­ence, even one per­son is enough to achieve real gains in a test­ing pro­gram. Some enter­prise global power users of Adobe Tar­get, for exam­ple, have one to three mem­bers on their opti­miza­tion teams and are gen­er­at­ing hun­dreds of mil­lions of dol­lars in rev­enue lift. The main rea­son why these pro­grams can func­tion with a small team is that the essen­tial opti­miza­tion steps in the pro­gram are con­ducted within a well-established work­flow, lend­ing itself to pre­dictabil­ity from an exec­u­tive expec­ta­tions stand­point as well as a report­ing per­spec­tive on when results will be deliv­ered. Do you have an exec­u­tive spon­sor? Do you have the ever-important project man­ager respon­si­ble for man­ag­ing resources and sched­ul­ing? Is there an avail­able cre­ative resource or can the opti­miza­tion man­ager make the desired con­tent edit­ing within the solu­tion as needed? This orga­nized test­ing work­flow can be con­structed and man­aged like an assem­bly line, in which many tests are in vary­ing stages of devel­op­ment, exe­cu­tion, and analy­ses. Our built-in safe­guards for con­cur­rent test­ing are there to help main­tain accu­racy in report­ing and scal­a­bil­ity of test­ing activ­i­ties on a sin­gle page or across your dig­i­tal channels.

Another pit­fall I often see is that a pro­gram will equate the num­ber of tests cur­rently being run as the lit­mus test for suc­cess and momen­tum in its pro­gram. “The more data we receive from test­ing, the more accu­rate our tar­get­ing and per­son­al­iza­tion strat­egy will be.” Sixty poorly exe­cuted, sim­ple tests with lim­ited data to fil­ter reports can give you less data than one strate­gi­cally con­structed test with access to rich seg­men­ta­tion and suc­cess met­rics for fil­ter­ing and eval­u­at­ing suc­cess. A well-developed, pro­gres­sive test­ing and opti­miza­tion strat­egy with focused activ­i­ties that have a clear objec­tive and mea­sure­ment of suc­cess is tan­ta­mount to the effec­tive­ness of your imme­di­ate return on invest­ment (ROI) as well as growth in terms of sophis­ti­ca­tion and pro­gram expan­sion. Effec­tive eval­u­a­tion of the results and informed com­par­i­son of the rel­a­tive influ­ence of key seg­ments and com­pound vari­ables can shine a light on the real per­son­al­iza­tion oppor­tu­ni­ties avail­able to you, as well as where your focus is not valuable.

There is also a reli­able strate­gic method­ol­ogy for eval­u­at­ing the most valu­able ele­ments to tar­get at each stage of the customer’s jour­ney: begin­ning with what is absolutely required on a given page or stage in the cus­tomer jour­ney and what is not, next eval­u­at­ing the essen­tial func­tions of each ele­ment, and then div­ing in to under­stand the rel­a­tive influ­ence of the ele­ments within cus­tomer con­tent. This method­ol­ogy iden­ti­fies the most rel­e­vant expe­ri­ences for dis­tinct seg­ments or indi­vid­u­als of your pop­u­la­tion using man­ual or (more likely) auto­mated capa­bil­i­ties for indi­vid­ual per­son­al­iza­tion based on iden­ti­fy­ing the most pre­dic­tive vari­ables within rich pro­file data. The suc­cess of this method­ol­ogy has moti­vated Adobe to embed more guid­ance, safe­guards, and best prac­tices within our new visual lin­ear work­flow, and to pro­vide a strate­gic advi­sor to assist with ideation and strat­egy devel­op­ment when needed.

Hav­ing well-defined roles, a reli­able work­flow, and a proven, strate­gic method­ol­ogy are the right ingre­di­ents for expe­ri­enc­ing true momen­tum in your opti­miza­tion pro­gram and see­ing imme­di­ate rev­enue lift that you can evan­ge­lize along with insights on refine­ment steps to focus on. Match this with the ease of use, gov­er­nance, breadth of man­ual and auto­mated capa­bil­i­ties, and syn­chro­nized pro­file data pro­vided within the Adobe Tar­get solu­tion, and I imag­ine I’ll be rack­ing up some seri­ous mileage rewards hear­ing about the ongo­ing suc­cess that more and more com­pa­nies are see­ing with test­ing and per­son­al­iza­tion across their dig­i­tal channels.

0 comments