Shop­per Re­search

Point of Purchase - - CONTENTS - By Liz Craw­ford, Se­nior In­dus­try An­a­lyst

The fol­low­ing is the 1st in­stall­ment in a six-part se­ries ex­am­in­ing best prac­tices for the mea­sure­ment of shop­per mar­ket­ing. This ar­ti­cle looks at ef­fec­tive ways to ra­tio­nal­ize the in­vest­ment. Sub­se­quent ar­ti­cles will cover the mea­sure­ment of shop­per be­hav­ior and brand impact, ef­fec­tive in­te­gra­tion prac­tices, re­tail col­lab­o­ra­tion and di­rec­tions for the fu­ture.

PART 1: Ra­tio­nal­iz­ing the In­vest­ment

Shop­per mar­ket­ing prom­ises big things. It prom­ises to fuel brand de­mand and in­flu­ence shop­per be­hav­ior while also sell­ing more prod­uct in-store. Brand teams like it so much that, ac­cord­ing to Shop­per Mar­ket­ing’s an­nual Trends Sur­vey, over half plan on in­creas­ing their shop­per mar­ket­ing bud­gets. Some of this in­creased fund­ing will be si­phoned away from above-the­line ef­forts. Yet, de­spite the grow­ing en­thu­si­asm for t his ap­proach, most brands still don’t un­der­stand ex­actly what they are get­ting for their money.

Ex­ec­u­tive Sum­mary

When the ba­sic unit of suc­cess is a dol­lar, a shop­per mar­ket­ing pro­gram’s other achieve­ments can be un­der­val­ued or for­got­ten en­tirely. Avoid the “mea­sure­ment trap” wher­ever pos­si­ble by track­ing met­rics other than sales.

The best in­di­ca­tor of suc­cess is what ev­ery­one agrees it should be. Get­ting agree­ment on suc­cess mea­sure­ments up­front is cru­cial to achiev­ing goals and build­ing bench­marks.

The level of spend­ing on a pro­gram should dic­tate the scope of most mea­sure­ment plans. The greater the spend, the more so­phis­ti­cated the met­rics.

In gen­eral, there are three tiers of mea­sure­ment by spend level:

Ba­sic sales Shop­per be­hav­ior Brand be­liefs and at­ti­tudes

The big­ger and more ex­ten­sively mea­sured pro­grams can serve as tu­to­ri­als for the rest of the or­ga­ni­za­tion, es­pe­cially when sys­tem­at­i­cally shared be­fore the next plan­ning cy­cle.

Avoid the im­pulse to com­pare and de­spair: the en­tire in­dus­try is on a learn­ing curve. Keep learn­ing and im­prov­ing.

of for­eign brands and in some cases are more trust­ing of for­eign/global brands. In China, nine out of 10 con­sumers say Shop­per mar­ket­ing has been likened to quan­tum me­chan­ics, the world of in­nites­i­mally small things. It is like mi­cro-mar­ket­ing on a mass scale. On the other hand, tra­di­tional mar­ket­ing is like New­to­nian physics, which takes a broader per­spec­tive on the world. Given th­ese dif­fer­ences, older re­search method­olo­gies (es­pe­cially broad-scale, na­tional ones) don’t cap­ture the en­tire impact of small­er­pen­e­tra­tion shop­per pro­grams.

Fur­ther­more, shop­per mar­ket­ing of­ten de­mands a broader, brand-ag­nos­tic mind­set that re­quires dif­fer­ent suc­cess mea­sures be­yond num­ber of cases sold – brand-port­fo­lio lift, out-of-store be­hav­ior changes, cat­e­gory lift and trip fre­quency, among oth­ers – that tra­di­tion­ally haven’t been part of the arse­nal. That makes com­par­ing the fi­nan­cial re­turns of shop­per mar­ket­ing ef­forts to those from tra­di­tional ac­tiv­ity very difcult.

Through­out the in­dus­try, mar­keters adopt­ing a shop­per strat­egy have been grap­pling with th­ese ques­tions, “How do I ra­tio­nal­ize my spend­ing, above and be­low the line? What am I get­ting for my in­vest­ment – re­ally? What should be the key met­rics across pro­grams?”

To un­der­stand th­ese is­sues and their po­ten­tial so­lu­tions in greater depth, Shop­per Mar­ket­ing in­ter­viewed dozens of ex­ec­u­tives across the in­dus­try, from re­tail­ers to re­searchers, agen­cies to brands. The result is a se­ries of six ar­ti­cles re­port­ing on best prac­tices in mea­sure­ment and the di­rec­tion re­search may be head­ing with ad­vances in tech­nol­ogy.

Key Is­sues and Hur­dles

Shop­per mar­keters are look­ing to ra­tio­nal­ize in­vest­ments as bud­gets shift from above to be­low the line. For most brands, there is a sin­gle pot of money for the to­tal­ity of mar­ket­ing ef­forts; it is a zero­sum game at the begin­ning of each ning cy­cle. But a por­tion of the money that used to be spent gain­ing im­pres­sions through mass me­dia ad­ver­tis­ing is now spent on shop­per initiatives. The ques­tion for many is, “How do I jus­tify this spend shift?”

Be­cause shop­per mar­ket­ing is more closely aligned with the pur­chase than tra­di­tional me­dia, it stands to rea­son that mea­sur­ing its impact should be easy. Or at least eas­ier. But this hasn’t been the case. Shop­per mar­keters have been stum­bling over var­i­ous ob­sta­cles in their quest to mea­sure impact. There seem to be four ma­jor stum­bling blocks, re­cur­ring across cat­e­gories and chan­nels.

1. The Mea­sure­ment Trap

Thirty years ago, pro­mo­tions were one-off events that were fairly sim­ple to mea­sure. A brand would need to move a cer­tain num­ber of cases to pay off an in­vest­ment in a dened pe­riod. This con­cept was easy to grasp. The mea­sure­ment for suc­cess was one-di­men­sional: sales vol­ume. But if you look at mea­sure­ment through the shop­per-mar­ket­ing lens, it be­comes more com­plex, be­cause shop­per mar­ket­ing is an ef­fort to pro­vide long-term rel­e­vance by de­liv­er­ing holis­tic so­lu­tions to shop­per needs – thereby build­ing the brand and chang­ing be­hav­ior in the process.

So what is the “mea­sure­ment trap”? De­fault­ing to sales met­rics, to the ex­clu­sion of other mea­sures. Tracey Doucette, se­nior vice pres­i­dent, cus­tomer strat­egy, eld and shop­per mar­ket­ing, at Pep­siCo, warns, “It is tempt­ing to use the met­ric we can eas­ily get, rather than mea­sure the ‘real’ ob­jec­tive of the pro­gram.” But the best shop­per mar­ket­ing pro­grams have ob­jec­tives that ex­tend well be­yond sales into shop­per be­hav­ior and at­ti­tudes. If only sales are mea­sured, the full yield of shop­per pro­gram­ming is ob­scured.

2.What Are We Mea­sur­ing Any­way?

In an at­tempt to cap­ture a fuller pic­ture of pro­gram per­for­mance, other acronymed suc­cess mea­sures have emerged, in­clud­ing: Re­turn on Ob­jec­tives (ROO), Re­turn on Mar­ket­ing Ob­jec­tives (ROMO), and Re­turn on Re­la­tion­ship (ROR), among oth­ers. ROO or ROMO usu­ally refers to mea­sur­ing shifts in shop­per be­hav­ior over time (be­yond the pro­mo­tional pe­riod), or to shifts in brand at­ti­tudes. Th­ese changes can be de­ter­mined quan­ti­ta­tively, al­beit at sig­ni­cant cost. The ROR is not of­ten a quan­ti­ta­tive mea­sure, but a con­cept that at­tempts to cap­ture the pos­i­tive impact of a shop­per pro­gram on the re­la­tion­ship be­tween a man­u­fac­turer and a re­tailer.

How­ever sig­ni­cant th­ese achieve­ments may be, they can be un­der­val­ued or for­got­ten en­tirely when the unit of suc­cess is a dol­lar, and only a dol­lar.

Ef­fec­tive pro­grams seek to change be­hav­ior well be­yond the pro­mo­tional pe­riod, ex­ert­ing a last­ing in­u­ence on shop­per habits. For ex­am­ple, the Kraft iFood As­sis­tant mo­bile app sug­gests meal so­lu­tions that bun­dle prod­ucts to solve a shop­per’s “What’s for din­ner?” dilemma. The app is also geared to meet busi­ness ob­jec­tives: driv­ing sales for Kraft prod­ucts, and in­creas­ing bas­ket ring for re­tail­ers. [Note: While some in­dus­try pro­fes­sion­als re­strict “shop­per mar­ket­ing” to ac­tiv­ity re­lated to par­tic­u­lar store en­vi­ron­ments (ei­ther lit­er­ally or through col­lab­o­ra­tive out-of-store pro­grams), gen­eral con­sen­sus denes it as any ac­tiv­ity that pushes a shop­per along the path to pur­chase, the in­ter­pre­ta­tion that will be used in this se­ries.] The in­dus­try is seek­ing to un­der­stand how th­ese kinds of shifts in shop­per be­hav­ior can be cap­tured in terms of met­rics, and then com­pared with re­sults from other pro­grams.

Shop­per mar­ket­ing can also in­flu­ence brand be­liefs and at­ti­tudes. Think of the pos­i­tive impact of the pink rib­bon on Yo­plait’s brand per­cep­tion, or the halo that Camp­bell Soup’s La­bels for Ed­u­ca­tion pro­gram gives to par­tic­i­pat­ing brands. In addition to driv­ing sales, th­ese pro­grams leave last­ing im­pres­sions on brand eq­uity. (This can hold true for both the brand and its re­tail part­ners.) But how is the brand eq­uity impact cap­tured here? And how do we gauge any ef­fect on the man­u­fac­tur­erre­tailer re­la­tion­ship?

To add yet an­other layer of com­plex­ity, stake­hold­ers have dif­fer­ing agen­das and de­n­i­tions of suc­cess. Pre­de­ter­mined met­rics need to reect the ob­jec­tives of the pro­gram, es­pe­cially when mul­ti­ple ob­jec­tives are in play. Without ex­cep­tion, ev­ery­one in­ter­viewed for this se­ries (agency, re­searcher, brand and re­tailer) as­serted that stake­hold­ers need to agree on the mark­ers of suc­cess, from the out­set, in or­der for suc­cess to be achieved. Pro­grams are so di­verse that it can be­come chal­leng­ing to com­pare one to an­other in “ap­ples to ap­ples” fash­ion. There­fore, it is im­per­a­tive to reach in­ter­nal con­sen­sus on suc­cess met­rics prior to pro­gram ex­e­cu­tion.

The chart on page xx iden­ties three buck­ets of shop­per-cen­tric mea­sure­ment: sales trans­ac­tion data, be­hav­iors, and brand at­ti­tudes/be­liefs. Man­u­fac­tur­ers and re­tail­ers each have goals in th­ese buck­ets, but they can dif­fer. In­cor­po­rat­ing both sets of goals into pro­gram­ming has be­come crit­i­cal to suc­cess. Deb­o­rah Han­nah, shop­per mar­ket­ing direc­tor at Star­bucks Cof­fee Co., says, “A self-serv­ing brand goal doesn’t cut it in shop­per – the shop­per ob­jec­tive needs to in­cor­po­rate re­tailer ob­jec­tives, such as share of wal­let, mar­gin growth, cat­e­gory growth, and bas­ket size.” (The task of rec­on­cil­ing ob­jec­tives and manag­ing the an­a­lyt­ics process with re­tail part­ners will be ad­dressed in ar­ti­cle ve, Col­lab­o­rat­ing with Re­tail­ers.)

Col­lab­o­rat­ing with the re­tailer on ob­jec­tives can bring its own re­wards. The chart also in­cludes a fourth bucket, the pro­gram’s impact on the brand-re­tailer re­la­tion­ship. While this is rarely a “hard” met­ric, each stake­holder hopes to re­ceive some benet, or re­turn, in this area as well.

3. The Du­bi­ous Prota­bil­ity of Mea­sure­ment

In­creas­ingly so­phis­ti­cated mar­ket­ing strate­gies have de­manded in­creas­ingly so­phis­ti­cated met­rics. And, for the most part, th­ese are avail­able. From in-aisle video mon­i­tor­ing, to track­ing in-store trafc pat­terns and eye move­ments, to link­ing on­line ex­po­sures to bricks-and­mor­tar pur­chases, shop­pers are be­ing tracked, tagged, asked and ob­served. Method­olo­gies abound. Why don’t re­sults?

The an­swer is that fund­ing state-of-theart met­rics tools would wipe out prota­bil­ity for many smaller pro­grams. It doesn’t make too much sense to spend $40,000 to mea­sure a $150,000 pro­gram.

As a prac­ti­cal mat­ter, then, not all shop­per mar­ket­ing pro­grams can – or need to be – mea­sured to the fullest. Ac­cord­ing to in­dus­try pro­fes­sion­als from both sides of the ta­ble, it seems that best prac­tices are tiers of mea­sure­ment com­men­su­rate with spend­ing. That is, the big­ger the bud­get, the more ex­ten­sive the per­for­mance met­rics. The size of the pro­gram sets the level of ex­pec­ta­tion for the met­rics; this makes sense from prota­bil­ity per­spec­tive, be­cause pro­gram­sef­fec­tively “buy” their own met­rics.

Fur­ther­more, the big­ger and more ex­ten­sively mea­sured pro­grams can serve as tu­to­ri­als for the rest of the or­ga­ni­za­tion, es­pe­cially when th­ese are sys­tem­at­i­cally shared be­fore the next plan­ning cy­cle.

Stick­ing with the three buck­ets as a sim­ple way of pars­ing met­rics, the scope of the pro­gram roughly par­al­lels the met­rics to be ob­tained (see chart on page xx). Vir­tu­ally all pro­grams, even one-off tac­ti­cal ef­forts, are ex­am­ined from a sim­ple lift stand­point us­ing sales data. Lift met­rics are then com­pared to his­tor­i­cal norms and bench­marks for that cat­e­gory.

A few larger, for­ward - lean­ing man­u­fac­tur­ers have im­proved upon sim­ple lift anal­y­sis to em­ploy sales data as in­puts into cus­tom mar­ket­ing mix mod­els. (“Mar­ket­ing mix mod­el­ing” refers to the use of mul­ti­vari­ate re­gres­sion and other sta­tis­ti­cal tech­niques on sales data to gauge the impact of var­i­ous mar­ket­ing tac­tics.) While nearly all of the larger con­sumer pack­aged goods brands run analy­ses an­nu­ally, us­ing na­tional data as in­puts, the more pro­gres­sive brands are us­ing desk­top mod­els on a brand-by­brand, pro­gram-by-pro­gram ba­sis.

Pro­grams with greater scope and spend­ing may war­rant ob­tain­ing the next level of mea­sure­ment: shop­per be­hav­ior in­for­ma­tion. This can come from shop­per card data, when ex­am­ined lon­gi­tu­di­nally and across bas­kets. Other kinds of be­hav­ioral met­rics can come from re­search speci­cally set up for the task, in­clud­ing shop­ping cart track­ers, in-aisle video mon­i­tor­ing, eye-track­ing, vir­tual store test­ing, and oth­ers. (Ar­ti­cle two will dis­cuss a num­ber of th­ese meth­ods.)

The big­gest cam­paigns, of course, are most likely to earn the most ex­ten­sive anal­y­sis, which would in­clude brand at­ti­tude mea­sure­ment. While most brands ex­am­ine shop­per be­hav­ior (via panel data) and brand eq­uity (at­ti­tu­di­nal track­ing stud­ies) on a na­tional ba­sis, th­ese met­rics are not usu­ally bro­ken out by re­tailer or by pro­gram. How­ever, the largest shop­per mar­ket­ing pro­grams have the funds re­quired to ac­quire th­ese data on a one­off ba­sis.

4. Hu­man Na­ture

Nearly without ex­cep­tion, the ex­ec­u­tives in­ter­viewed for this se­ries were a bit apolo­getic about their self-per­ceived lack of rigor in mea­sur­ing per­for­mance. Within the walled gar­dens of their com­pa­nies, shop­per mar­keters lament the short­com­ings of their meth­ods, even when they are rel­a­tively mi­nor. Th­ese short­com­ings seem more acute when com­bined with the sneak­ing sus­pi­cion that “some­one” out there is do­ing a bet­ter job. It’s hu­man na­ture to put one’s ex­pec­ta­tions a lit­tle above re­al­ity and adopt the “grass is al­ways greener” method of com­par­i­son. But psy­chol­o­gist Robert Bringle was prob­a­bly right when he de­scribed envy as

a pos­i­tive mo­ti­va­tor that in­spires peo­ple to work harder.

It is not sur­pris­ing that the mar­ket­ing in­dus­try ex­pects great things from mea­sure­ment to­day. It has only been about a dozen years since the smart­phone was in­tro­duced. Most pro­fes­sion­als re­mem­ber a time be­fore smart­phones, a time be­fore cell­phones and, for some, even a time be­fore color TV. Tech­nol­ogy is ac­cel­er­at­ing our ca­pa­bil­i­ties and our ex­pec­ta­tions, too. Mar­keters now feel a com­pul­sion to track ev­ery brand in­ter­ac­tion with ev­ery shop­per – ac­cu­rately, cheaply and in real time. And this seems com­pletely plau­si­ble. In fact, some­thing sim­i­lar to this sce­nario may be in the ofng through new tech­nol­ogy.

In the mean­time, how­ever, it can be dis­ap­point­ing to be con­fronted with the re­al­i­ties of legacy mea­sure­ment sys­tems. But is this dis­ap­point­ment re­ally war­ranted? Tra­di­tional above-the- line ad­ver­tis­ing never re­ally de­liv­ered – or even promised – a direct re­turn on in­vest­ment. While re­la­tion­ships can be drawn, no met­ric has ever proved it. Shop­per mar­ket­ing is be­ing held to a higher stan­dard. Per­haps this is be­cause new tech­nol­ogy now tan­ta­lizes with the prom­ise of per­fect data.

“Mea­sure­ment has al­ways been a chal­lenge. Tra­di­tional mar­keters, es­pe­cially in brand ad­ver­tis­ing, have had to rely on at­ti­tu­di­nal shift met­rics, as there are many other in­u­ences that oc­cur be­tween the brand ad­ver­tise­ment and the pur­chase,” says Fred Bid­well, ex­ec­u­tive chair­man of JWT Ac­tion.

“With shop­per mar­ket­ing, we have more in­u­ence through­out the pur­chase fun­nel. While chal­lenges still ex­ist, new mea­sures al­low us to look at the full spec­trum of at­ti­tu­di­nal and be­hav­ioral shifts – in­clud­ing sales.”

New met­rics for suc­cess are be­ing dis­cov­ered as new paths to pur­chase are be­ing ex­plored. The jour­ney may be con­found­ing at times, but the out­come – more ef­fec­tive mea­sure­ment – is worth­while.

Tracey Doucette Sr. VP Cus­tomer Strat­egy, Field & Shop­per Mar­ket­ing, Pep­siCo

Deb­o­rah Han­nah Shop­per Mar­ket­ing Direc­tor Star­bucks Cof­fee Co.

Fred Bid­well Ex­ec­u­tive Chair­man JWT Ac­tion

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.