Data Management Platform: Issues & Recommendations

Analytics

Today, the DMP (Data Man­age­ment Plat­form) is every­where, from the projects of mar­ket­ing boards to the columns of the spe­cial­ist press, includ­ing the ser­vice offer­ings of agen­cies. In a world where data is gold, the DMP presents a crit­i­cal chal­lenge, com­bin­ing tech­nol­o­gy, mar­ket­ing, media and orga­ni­za­tion, and is now far beyond the buzz stage. More and more com­pa­nies, how­ev­er, are won­der­ing about the DMP: why embark on such a project? What about its ROI? What are the busi­ness chal­lenges jus­ti­fy­ing its imple­men­ta­tion? And how to make the most of it? Find answers, chal­lenges and rec­om­men­da­tions about DMP in this arti­cle.

What is a DMP?

Real con­duc­tor of your data, the pur­pose of a DMP is to uni­fy and cen­tral­ize all avail­able data about an indi­vid­ual, whether they are owned data (nav­i­ga­tion, mobile, social net­work­ing, offline, CRM) or third par­ty data. More con­crete­ly, a DMP is a soft­ware plat­form allow­ing com­pa­nies to col­lect and cen­tral­ize cus­tomers and leads data, and to reprocess them (by uni­fy­ing them, enrich­ing them, seg­ment­ing them, and steer­ing them), in order to acti­vate them on the dif­fer­ent chan­nels with which the DMP is con­nect­ed. The chal­lenge is then to obtain the most com­plete view pos­si­ble of my audi­ence, in order to be able to cut it into seg­ments, and com­mu­ni­cate effec­tive­ly with them.

The chal­lenges of a DMP

The chal­lenges of a DMP are mul­ti­ple, but we can iden­ti­fy two main ones:

1) Opti­mize media pur­chas­es: At present, the major­i­ty of adver­tis­ers decid­ing to invest in a DMP do so to opti­mize their media spend: 87% cit­ed specif­i­cal­ly their will­ing­ness to improve ad tar­get­ing, while 83 % quot­ed a need to adapt the pur­chase of media audi­ence (Source: Win­ter­ber­ry study for IAB). Cur­rent­ly, it is extreme­ly dif­fi­cult to deter­mine what the yield of media pur­chas­es is: a DMP pro­vides vis­i­bil­i­ty into the prof­itabil­i­ty and per­for­mance of media expo­sures.

2) Opti­mize pur­chase tun­nel: The sec­ond issue of a DMP is relat­ed to opti­miza­tion oppor­tu­ni­ties in the con­ver­sion rate on the web­site and mobile, as well as the pur­chase tun­nel: the col­lect­ed and aggre­gat­ed data allows for a detailed and com­pre­hen­sive pic­ture of its audi­ence, and thus to cus­tomize as much as pos­si­ble the expe­ri­ence in the dig­i­tal ecosys­tem.

Suc­ceed at a DMP project

Nev­er­the­less, the com­plex­i­ty of a DMP requires you to fol­low cer­tain rec­om­men­da­tions to ensure the suc­cess of its imple­men­ta­tion:

1) Start with a sim­ple sce­nario: It is essen­tial to avoid going too fast, too fast, and to wish to use all data sources from the start. Instead, we must accept to start using sim­ple sce­nar­ios and enrich, refine, and ramp up grad­u­al­ly.

Let’s take the exam­ple of one of our cus­tomers, a Euro­pean tele­phone & tele­vi­sion oper­a­tor. Their first use case was to dis­tin­guish its cus­tomers from its leads, by “on-board­ing” in the DMP its CRM data. Once this dis­tinc­tion made, each seg­ment was offered dif­fer­ent pur­chas­ing tun­nel but also spe­cif­ic adver­tise­ments (dis­play and search): the imple­men­ta­tion of this basic sce­nario led to a decrease of 18% in CPA (Cost per Acqui­si­tion) and a 45% aug­men­ta­tion of the Dis­play CTR (Click Through Rate). When we are spend­ing mil­lions of euros in media pur­chas­ing, it makes a huge dif­fer­ence in terms of busi­ness impact!

2) Inte­grate CRM data with audi­ence pro­files: Sec­ond­ly, it is also essen­tial not to lim­it our­selves to the data obtained from online cus­tomers, but rather to strive to inte­grate CRM data into the audi­ence pro­file. This allows to adapt the pur­chase tun­nels thanks to the exist­ing data, in order to get an even clear­er and even stronger vision.

If I take the exam­ple giv­en above, the cus­tomer in ques­tion was able to iden­ti­fy the most effec­tive and appro­pri­ate offers for its cus­tomers, but also for the leads he had touched, thus man­ag­ing to refine his mes­sage and per­for­mance.

Again, before using third-par­ty data, using data from the com­pa­ny made the dif­fer­ence: in less than 6 months, the client was “even” and all rev­enues gen­er­at­ed were then prof­it!

What about you, what is your expe­ri­ence in DMP? Feel free to con­tin­ue the dis­cus­sion and share your views on the sub­ject in the com­ments !


Analytics
Olivier Binisti

Posted on 07-29-2015


Join the discussion