An Important Reason for Retailers to Focus on Cross-channel Marketing

Retail mar­keters are focus­ing more on cross-chan­nel mar­ket­ing than ever before, espe­cial­ly mobile chan­nels. They are look­ing to opti­mise offers, per­son­alise mes­sag­ing, and pro­vide con­tex­tu­al con­tent through mul­ti­ple channels.

There is an impor­tant rea­son for mar­keters to take a cross-chan­nel mar­ket­ing approach: attri­bu­tion. In order to make effec­tive deci­sions around dig­i­tal mar­ket­ing ini­tia­tives, you must be able to jus­ti­fy them finan­cial­ly. To do that, you must be able to attribute the impact of dig­i­tal and offline touch­points upon the pur­chase to the cor­rect channels.

When retail mar­keters have been asked, about four in ten say they will imple­ment cross-chan­nel attri­bu­tion ini­tia­tives and test tools for chan­nels and con­tent. While that per­cent­age leads all mar­ket­ing meth­ods expect­ed to be imple­ment­ed in 2017, the num­ber should be higher.

Effec­tive attri­bu­tion is essen­tial to meet­ing con­sumer expec­ta­tions these days. Organ­i­sa­tions must know what they want and be able to cre­ate sat­is­fy­ing inter­ac­tions on the right chan­nels. To accom­plish this, they need a full view of their cus­tomers at a one-to-one scale, which attri­bu­tion helps to provide.

Attri­bu­tion mod­el­ing is used to assign cred­it to each mar­ket­ing touch­point a cus­tomer inter­acts with. One customer’s jour­ney might look like this:

Instead of analysing the per­for­mance of each touch­point and chan­nel on its own, cross-chan­nel attri­bu­tion analy­sis allows you to see the impact of a touch­point as it relates to the entire jour­ney. In oth­er words, you’ll be able to iden­ti­fy how many emails were opened before a click was made or whether the inter­ac­tion with the spon­sored ad led to a prod­uct page view.

Start by map­ping the cus­tomer journey

Con­sumers now expect a flu­id expe­ri­ence on their jour­ney with a retail brand. They are using mul­ti­ple devices to con­nect with your brand, but they expect a seam­less expe­ri­ence, no mat­ter where they wish to go. To pro­vide this flu­id expe­ri­ence, you should be mea­sur­ing which types of engage­ments, offers, or con­tent types are important.

Through­out your cus­tomers’ expe­ri­ences, you must be able to record each inter­ac­tion and mea­sure the impact an inter­ac­tion had on future behav­iour. As you map the jour­ney, each inter­ac­tion should build toward some­thing, whether it is a pur­chase or a social share of your con­tent. By cal­cu­lat­ing attri­bu­tion, you can define how your team’s activ­i­ties led to that future action.

Bring cross-func­tion­al teams together

The key to dri­ving the best cus­tomer expe­ri­ence is to break down the silos in your organ­i­sa­tion. Tra­di­tion­al mul­ti­chan­nel approach­es have been built in silos, each with its own man­age­ment, opti­mi­sa­tion, analy­sis, and report­ing structures.

For­rester Research has found that 62 per­cent of firms “incent and mea­sure only the effec­tive­ness of a sin­gle cam­paign or touch strat­e­gy.” And yet 67 per­cent of mar­keters indi­cate that attri­bu­tion is “high­ly valu­able and helps make smarter mar­ket­ing and media deci­sions.” Ade­quate­ly attribut­ing con­ver­sions will be impos­si­ble if chan­nels are man­aged in silos.

All teams must be aligned with the cus­tomer. Per­son­al­is­ing your mes­sag­ing, and engag­ing, con­vinc­ing, and retain­ing cus­tomers. These achieve­ments come from being able to mea­sure, respond, report, and share cus­tomer behaviour.

Inte­grate your tools

The next step is to have the tools that help you com­plete effec­tive attri­bu­tion analy­sis then push out opti­mised con­tent through all your chan­nels and plat­forms. You need to get audi­ence insights, opti­mise media, col­lab­o­rate on and man­age assets, tar­get and per­son­alise con­tent, and more. Stitch­ing togeth­er dif­fer­ent sys­tems could work, but an inte­grat­ed toolset improves tim­ing and deci­sion-mak­ing remarkably.

Data has to be shared across your enter­prise. Your sys­tems must be able to push out data to stake­hold­ers quick­ly so teams can take prompt and deci­sive action. If your sys­tems are not inte­grat­ed tight­ly, you may lose valu­able oppor­tu­ni­ties to under­stand cross-chan­nel attribution.

Ulti­mate­ly, the tools you rely on should allow you to not only con­duct attri­bu­tion analy­sis, but be able to react imme­di­ate­ly by automat­ing con­tent selec­tion and deliv­ery. This takes a lev­el of inte­gra­tion that few dig­i­tal mar­ket­ing ven­dors can provide.

What steps can you take to improve attribution?

You have to take two impor­tant steps to real­ly excel at cross-chan­nel attribution:

The first require­ments is to align peo­ple and process to become data-dri­ven. There is no more room for unver­i­fied claims of mar­ket­ing success—or lack there­of. Data should per­me­ate every step your mar­ket­ing organ­i­sa­tion takes to improve results. Again, this takes the right set of tools, but also a col­lab­o­ra­tive mind­set through­out the entire enterprise.

Sec­ond, your organ­i­sa­tion must take the painful but reward­ing step of becom­ing an expe­ri­ence busi­ness. It has to start at the top by get­ting exec­u­tives to bring focus into dri­ving flu­id, respon­sive expe­ri­ence that leave cus­tomers want­i­ng more. Teams should align fore­casts, bud­gets, tools, and objec­tives so all chan­nels under­stand and con­tribute to the entire jour­ney. Give teams incen­tives to collaborate.

If you think of the emerg­ing arti­fi­cial intel­li­gence mod­els, they are built on attri­bu­tion. Using pre­dic­tive ana­lyt­ics as the engine, machine learn­ing analy­ses the data and responds appro­pri­ate­ly. Data is col­lect­ed and mea­sured to antic­i­pate future behav­iour based on past activ­i­ty. This is using attri­bu­tion to the fullest.

What is inter­est­ing to note when one thinks of cross-chan­nel attri­bu­tion is that For­rester has found mar­keters using a vari­ety of attri­bu­tion mod­els to improve mar­ket­ing. While a lit­tle over one-third of respon­dents in its 2014 sur­vey indi­cat­ed they use a first-inter­ac­tion attri­bu­tion mod­el, 30 per­cent indi­cat­ed they use an “advanced, sta­tis­ti­cal approach, specif­i­cal­ly regres­sion-based mod­els” or a rules-based mod­el, “specif­i­cal­ly a self-defined weight­ed mod­el.” Over 10 per­cent cite using game the­o­ry models.

There may be no clear answer as to which attri­bu­tion method works best, but with­out an attri­bu­tion strat­e­gy, it may be dif­fi­cult to cal­cu­late ROI effec­tive­ly, putting mar­ket­ing ini­tia­tives at risk.