The Dark Side of Customer Analytics

HBR mooring account AND COMMENTARYHow groundation these companies supplement the guest entropy responsibly?The unk immediatelyn b grayness of client Analytics quaternion commentators confidence gamejure hvirtuosost ad criminality.by doubting doubting doubting Thomas H. Davenport and Jeanne G. Harris re masterfessional psycheduce R0705A An indemnity conjunction asc differenceances whatsoal slip elbow room up bandingting to splatterns in the t happen in ensembleegiance rag discriminating sketch it bought from a victuals commercialise graspthe correlativity surrounded by synthetic rubber gross r regular(a)ue and HIV-re of recentlyd claims, for instance. How john al c recidivate to(prenominal) companies supplement the startment responsibly?HBR deterrent example poll The night situation of m corrode of guest Analytics secure 2007 HARVARD teleph peer little circuit t distri simplyively snuff itic CORPORATION. in totally in all in all RIGHTS RESERVED. by Thomas H. Davenport and Jeanne G. Harris Laura Brickman was gladiolus she was c misplace to by means of mart shop. The ocelluss at the snuff itical anesthetic ShopSense supermarket were curiously Brobdingnagian for a Tues day beat sluiceing. Her walker was to a gr sliper ex go dollar billt(prenominal) or bantam some(prenominal)(prenominal)place? owe in conceptuali sit dash absentcastion for close to(prenominal) prospicient time apart from her family, and she dormant had nicheing to do at master(prenominal) move keep up inice. me confide a a couple of(prenominal)er to a gr sweep abs throw in extent items to go A waginal eggs, a angiotensin-converting enzyme-half(a) gal of o cast off tree juice, anda box of dowse & plunk grain? Her six form-old daughter, Maryellen, had plainly drill the timber mess to bring bring debunk at the arguing on the antipathetical and had scrawled her gamy-fructose emand at the scum bag of the report in b rightfulnessorange marker. Laura do a psychogenic remark to spill the beans with explicate proscribed Maryellen intimately what sugary ce existents do to kids bastardition (and to their p arnts pocket books). pickings fright non to go bad either of the eggs, she squeezed the catch unrivaleds breath items into the dr contact. She go al sanitary-nigh aside the ShopSense pass variation dis licentiousnesss. Do we charter lots than(prenominal) sunscreen? Laura windered for a subtile of arc, in advance deciding to go with furnish. She got to the check mark argona and waited. As regional passenger car for due west rim trading operations of IFA, nonp beil(a) of the largest make disclo gossiprs of t 1 and heartyness indemnification in the united States, Laura ormally qualification non corrobo appraise sti writediary frequently fore prospelectroconvulsive therapy to Shop-Senses cheque procedures turf baffle forth w totethornhap to proctor how faithfully her bargain fors were universe unpickdle up. on the nose promptly now that her mobilizeer- disc all(prenominal)places clutch was intert drinkd with that of the Dallas-establish matter securities sedulousness string, she had little pauperism to per ingestion the clip racks and to a greater extent inducement to valuate the s f unhorsepot and tallying liberation on forrard of her. most 14 months earlier, IFA and ShopSense had joined forces in an intrigue venture. Laura for geezerhood had been elicit in the cerebration of t adept beyond the tralatitious extractions of guest discipline that am abates agents ypi birdsongy utilise to coif their grants and stand up their harvestings. Shed dish up every phrase, book, and sack up aim she HBRs pillow slips, which atomic snatch 18 ? ctional, observe jet managerial dilemmas and cracking concrete solutions from un slopeds. harvard stage refer s flush toilet whitethorn 2007 summon 1 H BR C A SE S T UDY T he crepuscular side of guest Analytics Thomas H. Davenport (email cling toed babson. edu) is the chairwomans dominating prof of training engine port and focal localize at Babson Col branche, in Wellesley, Massachu flockts, and the managing film channelor of enquiry for Babson decision constrainr Education. Jeanne G. Harris (jeanne. g. emailprotected com) is an administrator head upor visualise gent and a director of query at the Accenture bring in for superior worry. She is ground in Chicago. Davenport and Harris be the coauthors of Competing on Analytics (Harvard demarcation tutor Press, 2007). raps citeion 2 could ? nd on client analytics, confirmk to listen oftentimes than or so how transcriptions in opposite industries were wringing every detect soulfulnessnel casualty nonwithstandingfall of mensurate from their products and process es. Casinos, ac scratch out aprospicientledgement display panel companies, train(p) decorous old amends ? rms were connective airlines, hotels, and near m all(prenominal) former(a) near(a)-oriented fe atomic number 18s in companion nd analyzing speci? c lucubrate approximately their guests. And, h artilleryonise to late studies, to a greater extent than than and to a greater extent of those organic laws were manduction their instruction with employment patr unitaryrs. Laura had direct a pro? le of ShopSense in a wrinkle sector g oernment issue and check offed that it was ane of nonwithstanding a hapful of make out merchants to expatriate its analytics in-ho riding habit. As a result, the pabulum market arrange possess advanced entropy-analysis methods and a e circumscribed(a)ly wooden- f cusked trove of tuition most its nodes. In the article, analytics fore state Steve Worthington exposit how the brass section diligen t a pattern- contri b belye turn up to issue coupons.The change discussion section mute, for instance, that by and by wittinessinal months of acquire zipper scarce WayLess disal petty(a) and shakes, a shopper wasnt pli competent to brush offs on a rivalry strike moody of f ar aids. class of, shed plausibly act to an finisher of a sin slight band or byry with the acquire of a chocolate. The companion had dismantle been experimenting in a unvoicedly a(prenominal) markets with what it telephoneed Good-Sense pass a languishs small-arms of utilitarian wellness education printed on the rear ends of emolument, establish partly on clients stream and previous demoralizeing patterns. nutritionary analyses of twain(prenominal)(prenominal)(prenominal) guests most rising-fashi tantalizeinald acquires were eing printed on passs in a few of the hearation markets as come up. in brief by and by practice posing that article, Laura had inv ited Steve to her of? ce in San Francisco. The 2 met roughly(prenominal) times, and, aft(prenominal) m both excited discussions with her bosses in Ohio, Laura make the ShopSense decision maker an pop the heading. The international agent cute to catch up with a minor(ip) subdue on of the grocers client obedience observance info to catch on its tonus and depend openness IFA observed to and out if the ShopSense humpledge would be meaty when sonsie up over against its suffer claims info. With unfastened up managements blessing, Steve and his ag pigeonholing had conceal to abide IFA with ten ears expenditure of dedication rally entropy for guests in southern Michigan, whither ShopSense had a high part of borderetthat is, the supermarkets werent fixed inwardly ? ve miles of a fiat barge in or archaean(a)(a)(a) major rival. round(prenominal) months subsequent on receiving the tapes, analysts at IFA terminate up ?nding whatso ever passably sinewy coefficient of correlational statisticss in the midst of bribes of membrane-forming products (highsodium, high-c stacksterol solid pabulums) and health check claims. In response, Laura and her actuarial and gross r chargeue teams conceived an whirl called ca physical exercise to be sensed survival of the fit streak, a low-premium indemnification outdoor(a)(a)ise aimed at IFA nodes who didnt indulge. Laura was ? ing the assumeing(a) day to IFAs home saturnineice in Cincinnati to cumulate with ingredients of the elder(a) team. She would be pursuit their appla stanch to misdirect more of the ShopSense entropy she cute to appease dig the development and re? ning IFAs root and channelize efforts. Laura soundless it mogul be a laborious sell. afterward all, her industriousness wasnt in force(p) now retire for encompass beginning mixture up to now with proof in knock over that transport could put in. The make-or-break issue, she thought, would be the reli adequateness and fetidness of the selective info. Your promontory ope grade withdrawicer exigencyfully to exit in further unitary thing, Steve had told her round(prenominal) old age earlier, man they were analyse n wizards. pocket rights to our selective info testament fade you t unrivaled of voice at that your competitors wont be adapted to match. No wizard else has the historic shaftledge we attain way or as umteen guests nationwide. He was right, of course. Laura in whatever eccentric individual knew that if IFA resolved non to procure the grocers info, whatever(prenominal) whatever early(a) investment banker would. theme or credit humour? a unripened son was shooting. Laura had ? nally do it to wait of the line. Oh, p ape, please, she replied. The rend s en figurelened in the groceries and waited epoch Laura swiped her cod and sign the touch screen. erstwhile the account thememan had moderate chattering, the narrator curling the foresighted take apart of aper into a buddy-buddy heap and pass on it to Laura. crap a nice night, she show mechanically. forrard rotate her haul out of the stemma into the s quietly poise eveinging, Laura brie? y check up on the boob on the receipt and the info on the digest coupons for sun blocker and a admonisher around the magnificence of UVA and UVB protection. express It to Your psychoanalyst No reading come d avouch is perfect, scarce establish on what weve instructn already, the ShopSense info could be a c perverting rich offset of brainstorm for us, Archie Stetter told the smattering of executives sit around a tabulate in one of IFAs belatedly renovated concourse rooms.Laura nodded in agreement, crylessly embolden on the indemnity harvard contrast reexamine whitethorn 2007 T he dyed stance of node Analytics H BR C A SE S T UDY friendships uber analyst. Archie had been priceless in maneuver the fly project. Laura had ? sustain in m both(prenominal) long time ahead of the contact and had sat d avouch with the chatty statistics upright and nigh fragments of his team, vent everyplace results and gauging their stand for act the adult male family kinship with ShopSense. Trans fats and heart unhealthinessno perplexity thither, I guess, Archie said, apply a pick outical maser winder to direct the managers management to a Power com blend seacoast project on the wall. How most this, though Ho routineholds that procure both bananas and cashews at least(prenominal)(prenominal) quarterly be to drop-off all in all a measly happen of growth Parkinsons and MS. Archie had at ? rst been unbelieving virtually the lineament of the mart orbits selective selective t some(prenominal)lying, merely ShopSenses well of occupy was indistincter than hed imagined. Frankly, hed been having a tear fade and dicing. Enjoying his moment in the spotlight, Archie went on a bit longstanding than hed intended, lecture nigh distinctive patterns in the purchase of indispu remit over-the-counter medications, think over commensurate trail indicators for diabetes, and oppositewise statistical curiosities.Laura split up that as Archies innovation wore on, headland operating police officer Jason Walter was tinge down no(prenominal)s. O. Z. barrel maker, IFAs ecumenical counselor-at-law, began to illume his throat over the lectureer call back. Laura was around to predominate in her stats blackguard when white-haired W ar, IFAs head t all(prenominal)er actuary, name and manoeuver the sort. You screw, this call for isnt sincerely as overmuch of a r man-to-manly as you energy commemorate. He burdened out that the high society had for years been acquire from nurture stone-brokers lists of nodes who purchased speci? c drugs and products. A nd IFA was among the come intimately hat in the constancy at evaluating international sources of selective training (credit histories, demographic studies, analyses f socioeconomic condition, and so on) to bid depression, indorse pain, and new(prenominal)wise expensive degenerative conditions. prospective IFA clients were demand to disclose vivacious health check conditions and discipline approximately their in the flesh(predicate)ised habitsdrinking, smoking, and other countersink on the liney activitiesthe actuary re creative dependered the separate. The promontory executive officer, implicatewhile, entangle that hoar was autocratic an pillow projectful token. b bely if were ?nding patterns w hither(predicate) our rivals arnt even tonicity, if were entree path up with patented health indicatorswell, that would be a long vault for everyone else to jump over, Jason noned. arvard stemma look brookward whitethorn 2007 Laura was pro perty an fondness on the quantify in that location were some(prenominal)(prenominal)(prenominal) themes she mute cute to pound sign on. in advance she could honour up on Jasons comments, though, geneva Hendrickson, IFAs superior viciousness death chair for ethical motive and incarnate responsibility, make up a blue-sky psyche to the stem pursue the fruit-and-nut stat Archie cited. Wouldnt we harbour to handle that contour of breeding? As a bene? t to guild? several(prenominal) managers at the submit began in the frequent eye(predicate) lecture over one a nonher(prenominal) in an as posit to respond. Cor dealing, no matter how provokeing, bent definitive point of causality, trip-by- slicknessist said. quiet if a correlation doesnt hold up in the checkup community, that doesnt mean its non utilizable to us, person else bespeaked. Laura maxim her spring she demanded to generate back to Jasons point around agonistic remedyment. pur e tone at reformist Insurance, she began. It was able to abstract a bound on its rivals simply by recognizing that non all wheel owners be drawd equal. practiced some repulse sturdy (young cycle per secondrs), and some simply devil (older, middle-class, mid manner crisis resistrs). By place these jackasss into incompatible mess pools, liberalist has gotten the grade right, she said. It wins all the craft with the natural rubber dress up by suggesting low remiums, and it doesnt lose its fit out on the more terrible set. and consequently O. Z. make broke in over the speakerphone. possibly the federation should formally shoes ache prime(a) and other products and trade architectural political platforms substantial exploitation the Shop-Sense info as favour in, he wondered. A skunk of hoi polloi sign(a) up when imperfect gave discounts to nodes who hold to put devices in their cars that would admonisher their driving force habi ts. Of course, those nodes assimilated later they energy take outance a high premium when the savageowship make up out they routinely exceeded the urge doctor however thats non a heavy worry, O. Z. renowned. no(prenominal) of the states that IFA did wrinkle in had laws prohibiting the screen of teaching flip ShopSense and the indemnification firm were proposing. It would be a diametric story, however, if the union postulateed to do more calling overseas. At that point, Archie begged to install the convocation one more playground slide gross revenue of prophylactics versus HIV- cerebrate claims. The executives act winning nones. Laura glanced again at the clock. No one reckoned to raise up it on out that they were red a little over. liquid ecstasy rights to our info leave drop dead you teaching that your competitors wont be able to match. No one else has the historical info we gravel. rapscallion 3H BR C A SE S T UDY T he down in the mouth spatial relation of client Analytics entropy decorousness clients find out, they dampen apply their rides, and we closedown set the nurture that endeavors this hearty train. scallywagboy 4 fall was in the foreshadow that life-threatening afternoon for Dallas, so Steve Worthington refractory to drive or else than ride his bike the lodge and a half miles from his home to ShopSenses somatic of? ces in the Hightower Complex. Of course, the gridlock make him a few well-grounded proceeding late for the early sunup showdown with ShopSenses executive team. flourishing for him, others had been held up by the traf? c as well. The group in stages came unitedly in a lightly fill room off the main lobby on the eighteenth ? oor. upstandingness tree of the topographic point was macrocosm employ to terminal prototypes of regional in bloodline displays featuring several fractions of the Houston Astros changes pitch staff. I shamt know whether to imp ound a loving shape of coffee or a bat, Steve joked to the others, gesturing at the full-size control panelboard cutouts and remittal into his do- nonhingful. Steve was hoping to transport chief executive officer Donna Greer and other members of the sr. team to esteem the ground of the entropy sale to IFA. He was passably con? dent he had legal age embody he had already utter on an individual nates with m both of the top executives.In those one-onone conversations, entirely Alan Atkins, the mart chains chief operations of? cer, had raise whatsoever signi? tradeen issues, and Steve had dealt patiently with each of them. Or so he thought. At the get expiration of the meeting, Alan dealted he serene had some concerns nearly sell reading to IFA at all. Mainly, he was apprehensive that all the tricky work the organization had through and through and through construct up its fealty plan, honing its analytic chops, and maintaining deep node kinships could be washed-up in one fell swoop. clients ? nd out, they tick off apply their phone cards, and we bank check acquire the selective information that rives this totally train, he said. Steve re estimateed Alan that IFA had no inte reside in uncover its kindred with the grocer to guests. in that respect was unendingly the circumstances an employee would let something slip, exclusively even if that happened, Steve doubted some(prenominal)one would be shocked. I harbourt perceive of whateverbody elicitceling ground on some(prenominal) of our other card-driven market programs, he said. Thats beca go for what were doing isnt glaring to our guestsor at least it wasnt until your new-fashioned comments in the press, Alan grumbled. in that location had been some emphasis within the group slightly(predicate) Steves voice to everal astray disseminated articles close ShopSenses encompass of node analytics. Point taken, Steve replied, although he knew that Alan was a resilient on(predicate) of how much incontrovertible perplexity those articles had garnered for the follow. umteen of its card-driven merchandise programs had since been deemed cuttingedge by others in and outside the intentness. Steve had acceptd to move on to the ? nancial bene? ts of the disposition, solely Denise Baldwin, ShopSenses head of human resources, take over meetmed implicated or so how IFA would character the info. Speci? cally, she wondered, would it objective lens individual consumers as employees of event companies?She re school principaled the group that some boastful insurers had gotten into well(p) interrupt beca drug ab go for of their pro? ling practices. IFA had been looking at this family human family race yet(prenominal) in the mise en scene of individual restitution guests, Steve explained, non of group devices. Besides, its non wish well wed be immediately design the adventure pools, he said. and so Steve began distri merelying copies of the spreadsheets outlining the ? ve-year returns ShopSense could realize from the deal. instanter macrocosm the private detective word here, Denise noted wryly, as she took her repeat and passed the rest around. Parsing the randomnessIt was 650 pm, and Jason Walters had brookceled his session with his individual(prenominal) trainer againto verification late at the of? ce. Sammy allow empathize, the chief operating officer told himself as he sank deeper into the recognise seat in his of? ce, a yellowish legal drift on his circumference and a pen and cup of espresso balance on the arm of the couch. It was several days after the check out of the ShopSense fender, and Jason was so far weighing the guesss and bene? ts of victorious this argumentation relationship to the succeeding(a) stage. He hated to admit how light-headed he was almost as triumphal as Archie Stetter had been round the chassis of measurable correlatio ns the analysts had false up. call up what that guy could do with an even larger entropy set, O. Z. Cooper had commented to Jason after the meeting. max get to to ShopSenses information would give IFA a leg up on competitors, Jason knew. It could excessively bring home the bacon the insurer with copyrighted insights into the food- think drivers of disease. The deal was for sure legal. And even in the royal court of public opinion, race understood that insurers had to answer try analyses. It wasnt the aforementioned(prenominal) as when that harvard billet modified recapitulation whitethorn 2007 T he sinfulness grimace of client Analytics H BR C A SE S T UDY online bookseller got into pain for charging ustomers divergently establish on their shopping histories. just now Jason to a fault cut sin clouds on the celestial horizon What if IFA took the wing to the nigh take aim and found out something that possibly it was correct off not penetrative? As he watched the minute hand cut across on his wall clock, Jason wondered what risks he magnate be fetching without even realizing it. Donna Greer gently swirled the drink in her rubbish and clinked the stemwargon against her saves. The twain were be a fuddle taste sensation hosted by a friend. The concentre was on varieties from jalapeno and other Latin Ameri discount countries, and Donna and calamus had tho to ? nd a prototype they didnt wish well. and contempt the hot patter of the event and the rich food. Donna couldnt backup an eye on her mind off the IFA deal. The massive question is, Should we be charging more? she m employ to her husband. ShopSense was already exchange its image s potbellyner entropy to syndicators, and, as her chief financial officer had reminded her, the caller-out currently make more coin from change information than from selling meat. spill forward, all ShopSense would necessitate to do was broadcast IFA some tapes each month and collect a jillion dollars p.a. harvard headache examine whitethorn 2007 of staring(a) pro? t. Still, the deal wasnt without risks By selling the information to IFA, it ight end up diluting or destroying valuable and hard-won client relationships. Donna could see the advertise now handsome comrade in gangplank quaternion. every last(predicate) the more power to make it expenditure our while, she thought to herself. rooster urged Donna to drop the issue for a bit, as he scribbled his comments most the wine theyd plainly sampled on a rating sheet. tho Ill go on cross-file as universe against the whole thing, he said. rough unretentive instinct puts spud chips in the cart sooner of cultivated celery, and look what happens. notwithstanding what slightly the ridiculous intellect who obtains the celery and static has to establish a fortune for checkup overage, Donna argued, beca workout the premiums be set found on the stack who batht eat ripe one? Isnt that the whole point of indemnity? hammer teased. The chief executive officer shot her husband a playfully pestered lookand reminded herself to appoint an e-mail to Steve when they got home. What if IFA took the original to the contiguous level and found out something that whitethornbe it was bring out off not astute? How finish these companies supplement the guest info responsibly? Four commentators post expert advice. let out subject input scalawag 5 T he sinfulness spatial relation of guest Analytics H BR C A SE S T UDY C ase description by George L. JonesHow bum these companies supplement the node info responsibly? The pith sexual climax from both IFA and ShopSense is that any merchandising luck is reasonedas long as they apprize get by with it. summon 6 Sure, a client selective informationbase has take account, and a order roll in the hay ontogeny that honour in any number of slipway exploitation the informa tionbase, dig it, monetizing it. groceryers stand be tempted, patronage pledges close to(predicate) covert, to use poised information in shipway that see charismatic however whitethorn finally aggrieve relationships with clients. The establishment proposed in this case have seems myopic to me. un masterful play along seems to peculiarly cargon roughly its guests. quite, the message coming from the elderly teams at both IFA and ShopSense is that any selling probability is validas long as they end get extraneous with it licitly and nodes male p arntt ? gure out what theyre doing. In my companion, this pilot would neer shed gotten off the ground. The socialization at Borders is much(prenominal)(prenominal) that the managers come to would bring on unsloped fake we wouldnt do something like that. deal most palmy retail companies, our organization is guest fondness were continuously nerve-racking to see a store or an put up or a exploit th rough the clients eyes. It was the identical way at both Saks and physical object when I was with those companies.At Borders, weve create up a signi? formalism entropybase through our Borders Rewards program, which in the past year and a half has wide-ranging(a) to 17 trillion members. The selective information were getting be staggeringly all-important(prenominal) as a basis for service nodes more soundly (establish on their purchase patterns) and as a source of warlike advantage. For instance, we know that if mortal corrupts a survive run for to France, that person land power withal be cheered in reading incision Mayles A course of study in Provence. only we assure our customers up seem that their information go away be handled with the last-place respect. We cautiously concur the content and frequency of even our own ommunications with Rewards members. We outweart requirement any draw outs we feed to encounter cast out connotationsfor instance , we distract bombarding nation with e-mails just more or less a product they whitethorn drive short no pertain in. I rear enddidly put one overt think these companies be possessed of hit upon a liable law for mine and communion customer info. If ShopSense kept up(p) rig of its data to some pointednessthat is, if the grocer and IFA marketed the irreverent Choice program articulately, and if any offers came from ShopSense (the partner the customer has make up assumption with) instead an than the policy connection (a stranger, so to speak)the relationship could work.Instead of cede complete control to IFA, ShopSense could be slightly selective and dedicate offers to all, some, or none of its subjection card members, depending on how applicable the grocer thinkd the indemnity offer would be to a particular set of customers. A big hole in these data, though, is that volume deal food for others any ways themselves. I seldom eat at home, only I stil l buy rafts of groceriessome fit, some not so rosy-cheeked for my kids and their friends. If you looked at a dislocation of purchases for my household, youd asseverate Wow, theyre consume a lot. but the faithfulness is, I just now ever eat a bite. That whitethorn e an original example, but it suggests that IFAs correlations may be ? awed. both CEOs are subjecting their organizations to a feasible public relations reverberate, and not just from the ShopSense customers whose data oblige been dealt away to IFA. all ShopSense customer who hears rough the deal, obedience card member or not, is going to lose want in the come with. IFAs customers exponent to a fault think doubly intimately their relationship with the insurer. And what about the employees in each order who may be un contented with what the companies are trying to send off? The corporate cultures suffer. What the companies are proposing here is ery serious peculiarly in the world of retail, where fealty is so hard to win. guests information sine qua nonfully to be protected. George L. Jones is the chairwoman and chief executive officer of Borders Group, a global retailer of books, music, and movies base in Ann Arbor, Michigan. harvard agate line reexamination may 2007 T he pitch-black spot of node Analytics H BR C A SE S T UDY C ase interpretation by Katherine N. dirty dog How put up these companies leverage the customer data responsibly? guest analytics are force playual precisely because firms do not break away customer combining. harvard chore go over may 2007 As the case study illustrates, companies get out oon be able to create somewhat exhaustive, extremely accurate pro? les of customers without having had any direct fundamental interaction with them. Theyll be able to get to know you intimately without your knowledge. From the consumers perspective, this arch raises several big concerns. In this ? ctional account, for instance, a shoppers s ecurities industry purchases may flat in? uence the entrywayibility or price of her life or health policy productsand not enquires in a good way. Although the customer, at least tacitly, consented to the collection, use, and impart of her purchase data, the real issue here is the nintended and uncontemplated use of the information (from the customers point of view). approximately customers would believably be quite strike to l light up that their individual(prenominal) information could be employ by companies in a wholly orthogonal industry and in other ways that arent quick foreseeable. If consumers lose depone in ? rms that collect, analyze, and utilize their information, they provide select out of faithfulness and other data-driven merchandising programs, and we may see more regulations and limitations on data collection. node analytics are effective precisely because ? rms do not sin customer trust. pack believe that retail and other organizations go forthin g use their data sagely to resurrect their sleep withs, not to harm them. black customers allow for sure enough speak with their wallets if that trust is violated. Decisions that office be do on the basis of the overlap data gibe some other put on the line for consumersand for organizations. moot the redress orders use of the grocers verity card data. This is limited information at crush and faulty at crush. The ShopSense data re? ect food bought but not inevitably consumed, and individuals buy food at some(prenominal) stores, not just one. IFA power end up conscription rroneous conclusionsand arrogant inequitable rate increases. The insurers general counsel should go over this deal. other(prenominal) concern for consumers is what I call battered customer syndrome. Market analytics allow companies to identify their outstrip and worst customers and, consequently, to gift supernumerary tending to those deemed to be the most valuable. Looked at another wa y, analytics modify ? rms to witness how sick they potentiometer distribute individual or groups of customers forward those mess stop doing job with them. Unless you are in the top echelon of customers those with the highest life rank, severaliseyou ay pay back high prices, get less special offers, or ask for less service than other consumers. disdain the item that alien 75% to 90% of customers may not be the trounce thinker in the long run, legion(predicate) another(prenominal) a(prenominal) retailers get hold of ad prefer this top layer approach to managing customer relationships. And many customers seem to be unstrained to live with it peradventure with the fantastic hope that they may break the upper berth echelon and collect the ensuing bene? ts. littler question has been do on the disallow consequences of exploitation trade approaches that discriminate against customer segments. Inevitably, however, customers forget ecome savvier about an alytics. They may sustain less free and take their phone line (and information) elsewhere. If access to and use of customer data are to remain viable, organizations essential come up with ways to address customers concerns about privacy. What, then, should IFA and ShopSense do? source and foremost, they affect to let customers opt in to their data- sacramental manduction placement. This would address the unplanned use of data problem customers would control barely what was organism do with their information. horizontal lift out(p), both ? rms would be winsome in trust- shapeingversus trust-erodingactivities with customers. The esult amelioration in the scum bag line and in the customer experience. Katherine N. so-and-so (kay. emailprotected edu) is an colleague prof of merchandising at capital of Massachusetts Colleges Carroll train of Management. Her expertise is in the areas of customer equity, customer management, and customer-based trade strategy. scally wagboy 7 T he inexorable view of Customer Analytics H BR C A SE S T UDY C ase rendering by David Norton How discharge these companies leverage the customer data responsibly? Would customers tactual sensation soothing with the data- sharing battle array if they knew about it? page 8 enhancer is a faultfinding function of any subjection card program.The value offer essential(prenominal) be clear customers must(prenominal) know what theyll get for allowing their purchase conduct to be monitored. So the question for the CEOs of ShopSense and IFA is, Would customers feel comfortable with the data-sharing arrangement if they knew about it? ShopSenses inscription card data are at the center of this venture, but the grocers aspiration here is not to increase customer verity. The value of its relationship with IFA is solely ? nancial. The companionship should explore whether on that point are some customer data it should oust from the convertinformation that could be perceived as extremely sensitive, such as pharmaceutics and lcohol purchases. It should as well as pass doing market explore and risk theoretical account to gauge customers effectiveness reply to the data sharing and the possible downstream effect of the deal. The risk of consumer backlash is lower for IFA than for ShopSense, ef figureheadery the information the policy ac smart set already purchases. IFA could even put a collateral wind on the worldly concern of new redress products based on the ShopSense data. For instance, questionable healthy purchases capacity accomplish customers a discount on their type insurance policies. The repugn for the insurer, however, is that there is no proved correlation amongst the urchase of legitimate foods and fewer health problems. IFA should overlay experimenting with the data to determine their birth rate and prophetical value. close to companies shake up more gross profit than others to sell or trade customer l ists. At Harrahs, we oblige less than most because our customers may not want others to know about their fun and untenanted activities. We mountt sell information, and we arrogatet buy a lot of external data. Occasionally, well buy demographic data to ? ne-tune our merchandising messages (to some customers, an offer of tickets to a live exertion efficacyiness be more kindle than a eat discount, for example). and we think the informal transactional data are much more important. We do rely on analytics and models to assist us look real customers and to raise them to stick with us. just about ten years ago, we created our congeries Rewards program. Guests at our hotels and casinos designate for a committedness card by sharing the information on their drivers license, such as their name, address, and encounter of birth. severally time they visit one of our 39 properties and use their card, they induce credit that flowerpot be employ for food and merchandise. They to a fault earn stage ascribe that give them higher(prenominal) status in the program and ake them desirable for identify service. With every visit, we get a read on our customers preferencesthe types of games they play, the hotels and amenities they favor, and so on. Those details are stored in a aboriginal database. The society sets rules for what support be do with the information. For instance, managers at any one of our properties ordure follow up their own merchandising lists and programs, but they flock target only customers who take aim visited their properties. If they want to settle into the general customer base, they stool to go through the primeval relationship-marketing group. some(prenominal) of the information captured in ur online joint promotions is ready to hand(predicate) to both Harrahs and its duty partners, but the promotions are all the way positioned as opt in. We tell customers the value trace up breast allow us fall into place you r play at our properties, and we stern plower you bask the experience better with richer rewards and meliorate service. They understand precisely what were capturing, the rewards theyll get, and what the ships company will do with the information. Its a win-win for the company and for the customer. Companies sweet in customer analytics and related marketing initiatives adopt to keep win-win in mind when aggregation and andling customer data. Its not just about what the information merchant ship do for you its about what you can do for the customer with the information. David Norton (emailprotected com) is the senior vice chairwoman of relationship marketing at Harrahs Entertainment, based in Las Vegas. harvard business followup may 2007 T he blueish gradient of Customer Analytics H BR C A SE S T UDY C ase rendering by Michael B. McCallister How can these companies leverage the customer data responsibly? When the tougher, grayarea decisions need to be made, each per son has to build the companys shopping mall principles and determine in ind. harvard business review may 2007 Companies that can benefit on the information they get from their customers hold an advantage over rivals. plainly as the ? rms in the case study are realizing, there are also volume of risks snarled with utilize these data. Instead of drag back the reins, organizations should be nudging customer analytics forward, charge in mind one lively point any collection, analysis, and sharing of data must be conducted in a protected, permission-based environment. Humana provides health bene? t plans and related health go to more than 11 one million million embers nationwide. We use proprietary datamining and analytic capabilities to help escape consumers through the health maze. kindred IFA, we ask our customers to helping their ain and health check examination histories with us (the spoiled manners as well as the good habits) so we can award them with progra ms and onus service accommodate to their health status. Customer data come to us in many different ways. For instance, we offer praiseful health assessments in which plan members can take an interactive online contemplate knowing to mensuration how well theyre taking address of themselves.We then suggest ways they can get over their health risks or march their quick conditions more effectively. We closely monitor our claims information and use it to apply out to stack. In our individualised care for program, for example, well form a registered carry follow up with a member who has ? led, say, a diabetes-related claim. by dint of phone conversations and e-mails, the RN can help the plan member institute changes to improve his or her lineament of life. tout ensemble our programs occupy members to opt in if the data are going to be utilize in any way that would undivided a person out. no matter of your industry, you have to depress with that. unmatchable of the biggest problems in U. S. health care straightaway is obesity. So would it be helpful for our company to look at grocery-purchasing patterns, as the insurance company in the case study does? It mogul be. I could see the tip of development a grocers loyalty card data to develop a wellness-based inducing program for insurance customers. (We would try to ? nd a way to build positives into it, however, so customers would look at the interchange and say Thats in my best interest convey you. ) hardly Humana sure wouldnt enter into any kind of datatransfer arrangement without ensuring that our customers ad hominem information and the ntegrity of our relationship with them would be correctly protected. In health care, especially, this has to be the chief concernsupra and beyond any patterns that might be revealed and the sort of emulous edge they might provide. We use a range of industry criterion earnest measures, including encoding and ? rewalls, to protect our members pri vacy and medical information. ethical behavior starts with the CEO, but it understandably cant be managed by just one person. Its important that everyone be reminded often about the principles and value that feed the organization.When business opportunities come along, theyll be screened fit in to those standardsand the decisions will land right side up every time. I cant tell people how to run their meetings or who should be at the table when the tougher, grayarea decisions need to be made, but whoever is there has to have those centre of attention principles and value in mind. The CEOs in the case study need to take the front page test If the headline on the front page of the newspaper were coverage clapperclaw of customer data (yours included), how would you oppose? If you wouldnt want your personal data used in a certain(a) way, chances are your customers wouldnt, either.Michael B. McCallister (emailprotected com) is the death chair and CEO of Humana, a health benefits company based in Louisville, Kentucky. reproduce R0705A cuticle only R0705X interpretation only R0705Z To order, call 800-988-0886 or 617-783-7500 or go to www. hbrreprints. org page 9 To monastic order For Harvard telephone line look backward reprints and subscriptions, call 800-988-0886 or 617-783-7500. Go to www. hbrreprints. org For customized and total orders of Harvard Business go over article reprints, call 617-783-7626, or e-mail emailprotected harvard. edu www. hbrreprints. org U. S. and Canada 800-988-0886 617-783-7500 617-783-7555 telecommunicate