What Is Personalization?

Adobe’s dig­i­tal mar­ket­ing def­i­n­i­tion of per­son­al­iza­tion is the right con­tent at the right time to the right per­son. We hear so often that senior lead­ers want to imple­ment per­son­al­iza­tion on their sites and they want it yes­ter­day. Gone are the days of “dumb” sites that offer the same con­tent to 100% of traf­fic. Today, mar­keters know that vis­i­tors and cus­tomers want expect a tai­lored expe­ri­ence to their own needs and interests.

How­ever when senior lead­ers want this kind of con­tent man­age­ment yes­ter­day there is a lot of pres­sure to show progress. Here are some strate­gies to get you started suc­cess­fully from the onset and avoid wast­ing time.

The Weak­est Strat­egy to Start

1) Super­fi­cial targeting

That’s me” imagery typ­i­cally is not worth the effort. What is “That’s me” imagery? These are images that try to match the imagery to the vis­i­tor. For exam­ple, you know that a large por­tion of your cus­tomers are 35–45 sin­gle moms so you decide to test show­ing ban­ners (same mes­sag­ing) with women and chil­dren to cre­ate a famil­iar­ity on the site to vis­i­tors who fall into this audi­ence. Results are typ­i­cally flat. The rea­son being folks already know what they look like. “That’s me” imagery doesn’t add any new per­sua­sion or address any hur­dles of a visitor.

What’s bet­ter? Show dif­fer­ent kinds of imagery or prod­ucts based on the vis­i­tor affin­ity. Tech­ni­cal vis­i­tors might respond bet­ter to prod­uct views while non­tech­ni­cal folks might respond to lifestyle or ben­e­fit imagery. Per­son­al­iz­ing imagery to reflect a desire or sug­gested prod­uct, a need, a hur­dle, or a ques­tion has been most effec­tive for our clients.

2) High-value audi­ence with low-value messaging

So many clients do lots of audi­ence research and invest tons of money in their data. They pour through their ana­lyt­ics, pur­chase audi­ence stud­ies, and con­tract for a Data Man­age­ment Plat­form. They iden­tify a high-value audi­ence and con­firm a respectable level of traf­fic. How­ever, many times after all that time and money is spent, these mar­keters don’t actu­ally have some­thing dif­fer­ent to say to these folks. For exam­ple, you find that vis­i­tors from organic Bing​.com have a lower engage­ment rate than organic Google​.com. There’s enough Bing traf­fic per month to war­rant address­ing it. Some mar­keters might change a head­line or per­son­al­ize “Hello Bing Vis­i­tors” but does that really address the fun­da­men­tal dif­fer­ences of the site’s Bing and Google visitors?

What’s bet­ter? Address the rea­sons why these vis­i­tors are dif­fer­ent than the rest of the audi­ences. In this exam­ple, the Bing search terms may be more branded than the typ­i­cal non-branded Google search terms. Bing vis­i­tors could be more famil­iar with your prod­ucts yet are taken to a page with too many prod­uct options that over­whelms them with too much vari­ety. A bet­ter strat­egy would be to eval­u­ate the Bing prod­uct searches and limit the prod­ucts shown to the most pop­u­lar five.

The Most Effec­tive Strat­egy to Start

1) Make user flow easier

We have seen con­sid­er­able suc­cess high­light­ing the prod­uct that was pre­vi­ously viewed or pre­vi­ously added to cart. By pro­vid­ing vis­i­tors a queue to quickly com­plete their pur­chase or revisit where they left off, vis­i­tors see less noise and are more likely to con­vert. Vis­i­tors who have pre­vi­ously seen con­tent may appre­ci­ate cut­ting that con­tent down to sim­plify the page. Tai­lor­ing the site based on past or in-session engage­ment can be the most effec­tive use of per­son­al­iza­tion because the tar­get­ing is based on declared inter­est of the vis­i­tor (through behav­ior) rather than inferred inter­est from pur­chased data. In this case, the vis­i­tor is show­ing you their inter­ests through the fil­ters they select, the prod­ucts and pages they view, and the clicks they make or don’t make. Act on it!

Next step? Go into your ana­lyt­ics and check out the vol­ume of vis­i­tors to a prod­uct detail or cart page and return to the home­page. What sort of traf­fic do you see? What’s the con­ver­sion rate of these folks, and is there oppor­tu­nity to improve that? What’s the aver­age order value? Are the orders a higher value for folks who move from the home­page directly to the cart? Or are the orders a higher value for folks who move from the home­page back into the site and even­tu­ally con­vert? Cre­ate a cam­paign to push vis­i­tors either directly to prod­uct detail or cart pages of the pre­vi­ously viewed product.

2) Address an audi­ence chal­lenge or desire

Vis­i­tors to your site have dif­fer­ent his­to­ries with your brand and a vari­ety of hang-ups to con­ver­sions. Iden­tify the pos­si­ble chal­lenges for each of your major seg­ments and address them full on. Show­case what mat­ters to them most.

For exam­ple, when apply­ing for a small busi­ness credit card, busi­ness own­ers who travel fre­quently may pri­or­i­tize travel rewards and lounge access over other fea­tures and high annual fees. House paint­ing busi­ness own­ers don’t travel often and may pri­or­i­tize sea­sonal cash flow sup­port and low annual fees over other ben­e­fits. By show­ing the vis­i­tor the ben­e­fit or prod­uct that is the most impor­tant to them, you are cut­ting down the noise and hur­dles to application.

Next step? Out­line the audi­ences that may respond best to each of your prod­ucts or ben­e­fits. Can you iden­tify these audi­ences through site behav­ior pur­chased data, CRM, or other means? If so, cre­ate these seg­ments and mon­i­tor the level of traf­fic per month. With rea­son­able lev­els of traf­fic, these seg­ments can show con­sid­er­able engage­ment and con­ver­sion rate lift through address­ing their indi­vid­ual priorities.

3) Give an offer to your high-value audience

To save your bud­get and pre­serve pres­tige, you prob­a­bly don’t want to dis­count to every­one. How­ever, you may want to entice a cer­tain por­tion of your vis­i­tors who you con­sider high value. For exam­ple, you may want to share a 10% coupon to vis­i­tors who are in DMAs near your brick-and-mortar stores. Through pre­vi­ous test­ing you know that vis­i­tors to your stores spend an aver­age of 25% more in store and so your ROI will be much higher for vis­i­tors near your loca­tions. Or you might want to move your ski mer­chan­dise so you tar­get a ski dis­count to vis­i­tors who have a ski affin­ity. Instead of shar­ing with all your vis­i­tors, you incen­tivize the right audi­ence with the high­est propen­sity for your offer.

Next step? Iden­tify the audi­ences with the biggest oppor­tu­nity. Con­sider geolo­ca­tion, prod­uct affin­ity, return ver­sus new, or CRM data. What offers are avail­able to share? Of course, con­firm with legal first that your cam­paign is com­pli­ant with your busi­ness and indus­try best prac­tices. While you should never dis­count based on race, color, reli­gion, gen­der, or national ori­gin there are many oppor­tu­ni­ties to incen­tivize anony­mous vis­i­tors based on affin­ity, loy­alty, or seen pre­vi­ous mar­ket­ing campaign.

Test first!

Test­ing is the only way to mea­sure the ROI of your efforts and evan­ge­lize this suc­cess at your orga­ni­za­tion. Good rule of thumb is to A/B test first and then imple­ment to 100% of your seg­ment in per­pe­tu­ity. Test­ing is valu­able as well as fun. Happy testing!

Brain­storm for Action

Make the user flow easier:

  • Are your prod­uct options over­whelm­ing? Pare down prod­uct rec­om­men­da­tions based on site refer­rer, site engage­ment, and pre­vi­ous visits.
  • Can groups of vis­i­tors skip some steps to make it eas­ier to con­vert? Can you remove con­tent for par­tic­u­lar vis­i­tors? Can you add more guided con­tent? Can you sug­gest loca­tion and com­pli­men­tary prod­ucts? Iden­tify which vis­i­tors need more guided help and which need fewer obsta­cles in their way.
  • How can you urge loyal cus­tomers and vis­i­tors who have started the fun­nel to get back into the fun­nel again quickly? Work with IT to build solu­tions to save cart and retar­get products.

Address an audi­ence chal­lenge or desire:

  • Do some geolo­ca­tions have issues with your page load time? Reduce con­tent and sim­plify page for these locations.
  • What are the expected cus­tomers for each of your prod­ucts? Are some for busi­ness and some for con­sumer users? Tech­ni­cal ver­sus non­tech­ni­cal? Con­sider your major visitor/customer seg­ments and iden­tify the con­tent and prod­ucts best for them.
  • How might your vis­i­tors cat­e­go­rize their options? By use case, by value prop, by price, user type? Con­sider which options might be most impor­tant to your major visitor/customer seg­ments and high­light them.
  • Can you up the ante depend­ing on the vis­i­tor sales cycle? Should you share dif­fer­ent mes­sag­ing or offers in his fourth visit from his first? Iden­tify offers for vis­i­tors who need that extra push.

Give an offer to your high-value audience:

  • Who are your best cus­tomers? Can you iden­tify these attrib­utes in your new vis­i­tors? Pur­chase data to expose your best prospects in your visitors.
  • Are you try­ing to move par­tic­u­lar mer­chan­dise or slow down sales for items in back­log? Incen­tive only the audi­ence most likely to respond.



Great post! I've been struggling to find a way to differentiate between bing an google organic visitors for some time now. The audience profiles look nearly identicle in nature with the exception of a heavier female demographic.