Accord­ing to an August 2013 Retail Touch Points arti­cle, 78 per­cent of con­sumers would buy from a retailer more fre­quently if they received per­son­al­ized offers—and 71 per­cent don’t believe retail­ers are effec­tively pro­vid­ing these offers.

Although more than 50 per­cent of online retail­ers agree that per­son­al­iza­tion is fun­da­men­tal to their online strat­egy, per­son­al­iza­tion tac­tics are often imple­mented with­out an estab­lished per­son­al­iza­tion strategy.

What Dri­ves Your Business

There is no blue­print for per­son­al­iza­tion strat­egy. It largely depends on your business’s mix of prod­ucts and cus­tomer behav­ior. For exam­ple, zip code demo­graph­ics may be highly use­ful to a travel site, where cer­tain zip codes are more likely to book busi­ness ver­sus fam­ily travel or high ticket ver­sus bud­get get­aways, or have res­i­dents requir­ing “sun get­aways” in the win­ter. Another site might find zip codes less useful.

The abil­ity to mine your own data for insights primes you for “know­ing thy cus­tomer.” Involve your data sci­en­tists as much as pos­si­ble in your strate­gic planning.

The Size of Your Catalog

The larger the cat­a­log, the more sat­is­fy­ing a per­son­al­ized expe­ri­ence is. A large cat­a­log reduces the effort required to find rel­e­vant prod­ucts and con­tent. A niche online retailer that sells a tight range of prod­ucts may be able to sat­isfy per­son­al­iza­tion with a less com­plex mer­chan­dis­ing and rec­om­men­da­tion strategy.

The Diver­sity of Your Customers

Do you sell inter­na­tion­ally? Do you cater to a mix of male and female, age groups, busi­ness and con­sumer, etc.? What cus­tomer per­sonas or seg­ments would make effec­tive per­son­al­iza­tion tar­gets? Which seg­ments do you wish to prioritize?

What You Know about Your Customers

Take inven­tory of the data sources you can use and apply to cus­tomer seg­ments. This may include, but is not lim­ited to

  • Account pro­file data
  • Loy­alty pro­gram data
  • Demo­graph­ics, zip code
  • Geolo­ca­tion, time zone
  • Device con­text
  • New ver­sus return­ing visitor
  • Customer/noncustomer
  • Refer­ral sources such as web­sites, affil­i­ates, mar­ket­ing cam­paigns, social net­works, and com­peti­tor sites
  • Key­word refer­ral data (when possible)
  • Session-based click­stream data
  • Site search input
  • Past search, browse, and pur­chase history
  • Wis­dom of the crowds”—people who take sim­i­lar jour­neys tend to do X, Y, or Z, and real-time analytics
  • Email pro­gram segments
  • Recent site or cart abandonment
  • Cart con­tents
  • Pur­chase history
  • Third-party and cross-channel data sources

Ulti­mately, as an online retailer, you want to rec­om­mend the most rel­e­vant prod­ucts and offers. For each insight you can glean about a vis­i­tor, ask your­self, “how does this matter?”

One approach is to start with spe­cific con­tent on your site, such as the home­page ban­ner. For exam­ple, a tele­com site may serve dif­fer­ent ban­ners depend­ing on whether a vis­i­tor is an exist­ing cus­tomer or likely with a competitor.

Exam­ple A:

welcome-back-1[2]

Exam­ple B:

new-customer-offer-2

Another approach is to work from the cat­e­gory or prod­uct level. How can you rank search/browse results based on what you know about the cus­tomer? Are cer­tain styles bet­ter sell­ers in Florida? Has the cus­tomer pre­vi­ously pur­chased from or searched your site for a cer­tain brand? Is the cus­tomer located in a coun­try to which cer­tain prod­ucts can’t be shipped? Do you know a customer’s size and can you fil­ter results accord­ingly? Can you serve a prod­uct page with cross-sells most likely to appeal to this cus­tomer based on his or her brows­ing his­tory or past purchases?

You may also work from the persona/segment level. When a cus­tomer from seg­ment X arrives at your site, his or her expe­ri­ence should be Y (on-page or through­out the site).

Your Cus­tomer Journeys

Don’t think just about what jour­neys are most com­mon, but also about how you can pro­vide a bet­ter jour­ney. Can you use per­son­al­iza­tion to elim­i­nate steps such as search/endless browse, main­tain scent-of-intent from refer­ring cam­paigns and domains, deliver tar­geted offers directly to the inbox, sup­port cross-channel and cross-touchpoint activ­i­ties, or encour­age larger bas­ket size? Can you iden­tify a cus­tomer at a point of just-about-to-abandon, and prompt with live chat, offer, or other content?

Your Tech­nol­ogy

Once you’ve assessed the range of the pos­si­ble, pri­or­i­tize your tac­tics based on expected impact. Then deter­mine if there is a gap between what’s desired and what’s pos­si­ble. Are you miss­ing data sources? Are you capa­ble of per­sist­ing data across touch­points? Does the required data live in silos, and how dif­fi­cult is it to knit the nec­es­sary data together?

Can your cur­rent tech­nol­ogy sup­port your strat­egy? What involve­ment do you need from IT and how soon can it be accom­plished? At what cost?

Many per­son­al­iza­tion solu­tions don’t inte­grate email, mobile, in-store, plas­tic loy­alty pro­grams, affil­i­ate data or even test­ing tools, but could greatly ben­e­fit from doing so.

For many, siloed data is only a per­ceived bar­rier to entry. The tech­nol­ogy to pull it all together already exists today in Adobe Mar­ket­ing Cloud. Adobe Mar­ket­ing Cloud already has the hooks to pull in data from any sys­tem, and feed it to your per­son­al­iza­tion, test­ing, and ana­lyt­ics, with min­i­mal involve­ment from IT.

The Out­come of Strategy

The point of your per­son­al­iza­tion strat­egy is not to do every­thing pos­si­ble, but to ensure that all data and use cases that could bet­ter serve rel­e­vant con­tent, prod­ucts, and offers are con­sid­ered, pri­or­i­tized, and planned for. A doc­u­mented plan also ensures that your mar­keters “stick to the script,” and employ the proper test­ing and val­i­da­tion processes to ensure your tac­tics are right for cus­tomers and your bot­tom line.

To learn more about Elas­tic Path, visit www​.elas​tic​path​.com.