What Are the Metrics That Matter?

We’re bom­bard­ed by data these days. Say I want to book a table for an impromp­tu meal with friends. A vis­it to the restaurant’s web­site pro­vides its address and hours, dis­plays its menu and prices, and pho­tographs of its sig­na­ture dish­es. They may cite the num­ber of tables, brag about the vari­ety of gin behind the bar, or offer chap­ter and verse about how the ingre­di­ents are sourced. With a bit more dig­ging, I can unearth pro­fes­sion­al reviews—or vis­it a site such as Trip Advi­sor, and read cus­tomers’ descrip­tions of their din­ing expe­ri­ences. But what if all I need on that par­tic­u­lar night is some­place equidis­tant from all our offices, near pub­lic trans­port? The rest becomes noise.

Data also comes at adver­tis­ers fast and furi­ous­ly, and busi­ness­es may strug­gle to deter­mine which suc­cess met­rics are the right ones to focus on when devel­op­ing an effec­tive, prof­itable dig­i­tal mar­ket­ing strat­e­gy. Tra­di­tion­al dig­i­tal media met­rics such as com­ple­tion rates, clicks, and cost per action don’t always eas­i­ly tie back to a brand’s bot­tom line. Nor do they nec­es­sar­i­ly give an actu­al mea­sure of the ad’s effectiveness—how do you know that your ad actu­al­ly caused that per­son to click?

To put that in per­spec­tive, between 2000 and 2009, the divorce rate in Maine and the state’s per capi­ta con­sump­tion of mar­garine were direct­ly cor­re­lat­ed, but it’s safe to assume that one did not cause the oth­er. The same goes for marketing—if some­one sees your ad and then buys your prod­uct, cor­re­lat­ed met­rics assume that the ad influ­enced the pur­chase, but was that per­son plan­ning to buy already?

This ambi­gu­i­ty makes it hard to allo­cate bud­gets appro­pri­ate­ly, let alone cre­ate a win­ning strat­e­gy. You’ll also be required to link dig­i­tal invest­ment to your organisation’s over­all busi­ness objec­tives (often to prove suc­cess to high­er-ups, or ensure ample bud­getary spend in the next fis­cal year). Thus mar­keters have quite a strug­gle ahead of them.

They know it, too. At the 2017 Digi­day Pro­gram­mat­ic Mar­ket­ing Sum­mit, mar­keters dis­cussed KPI over­load and their frus­tra­tion in try­ing to tie adver­tis­ing met­rics to busi­ness goals. A lack of qual­i­ty data, and an increas­ing­ly com­plex mar­ket­place were also cit­ed as con­cerns. What’s the solution?

Understand Different Data Sets

It has always been chal­leng­ing to link online mar­ket­ing spend with offline behav­iour, but thanks to the wealth of new­ly acces­si­ble data, those chal­lenges are wan­ing. The hur­dle now is sift­ing through all those met­rics to find the few that actu­al­ly have an impact on your brand’s unique goals. Not all met­rics are use­ful to all brands. There are dozens upon dozens of data points, and each comes with its own pros and cons. Some exam­ples include:

• SKU and prod­uct-lev­el met­rics. Com­pa­nies such as Nielsen, IRI and Kan­tar glean insights from pan­el data, loy­al­ty card data or point-of-sale infor­ma­tion in order to under­stand in-store pur­chas­ing behav­iour. This is great for brands seek­ing insights into how par­tic­u­lar prod­uct lines are per­form­ing across audi­ences and locations.
• Trans­ac­tion-lev­el data. Anonymized data on cus­tomer deb­it and cred­it card trans­ac­tions is avail­able from finan­cial insti­tu­tions. The main dif­fer­ence between trans­ac­tion data and SKU data is trans­ac­tion data can­not mea­sure sales at the prod­uct lev­el. This type of mea­sure­ment makes sense for retail and din­ing cat­e­gories. Due to pri­va­cy con­cerns, how­ev­er, access to this data is large­ly lim­it­ed in the UK and the EU.
• Loca­tion data. Loca­tion met­rics lever­age cell­phone data to pro­vide insights into where, when, and how cus­tomers move through the world. It’s ide­al for mea­sur­ing whether online mar­ket­ing drove real world store traf­fic. The loca­tion data space is crowd­ed, so it’s impor­tant to ass­es each provider and choose the right part­ner for your brand.
• Reach and fre­quen­cy (R/F) data. It’s impor­tant not to for­get about the mea­sure­ment tool that has been around since the Mad Men era. Lever­ag­ing part­ners like Nielsen or Atlas to mea­sure audi­ence and expo­sure, in con­junc­tion with action-based data, will help brands start to under­stand the effec­tive R/F lev­els to dri­ve busi­ness outcomes.

As the mar­ket demands more mea­sur­a­bil­i­ty, more solu­tions emerge dai­ly. It’s impor­tant for brands to seek out objec­tive part­ners when choos­ing a mea­sure­ment solu­tion, and under­stand that work­ing with­in a walled gar­den doesn’t always paint a full pic­ture. Inde­pen­dence is imper­a­tive for an unbi­ased view.

Determine What Matters for Your Brand

You’ll need to:

• Design exper­i­ments to reveal which media met­rics can be tied direct­ly to ROI—and which aren’t worth the effort.
• Ensure your exper­i­ments are analysing actu­al lift rather than cor­re­lat­ed met­rics, so you can be con­fi­dent that you are iden­ti­fy­ing the met­rics that actu­al­ly make a difference.
• Invest in the right omnichan­nel soft­ware that gives you a holis­tic view of all your mar­ket­ing efforts, includ­ing both online and offline sales.
• Find a good part­ner who’s well-versed in advanced media mea­sure­ment, and the var­i­ous data types and part­ners in the space. It’s a com­plex land­scape to tra­verse, so make sure you enlist help if you need it.

Remem­ber, it’s a crowd­ed field. If you’re not sure which data to use, or whom to trust to pro­vide it, rely on an expe­ri­enced third-par­ty such as Adobe Adver­tis­ing Cloud to guide the way. Hon­ing in on the right data—and find­ing the right part­ners to pro­vide it—will play a big role in your mar­ket­ing success.