Datahearts

Search
Close this search box.

Overcoming hurdles

Making better use of market and consumer data

How SMEs overcome hurdles in democratizing data science

Duality between data sources

Dis­cus­sions with con­su­mer and mar­ket insights teams often reveal a cer­tain dua­lity bet­ween insights from mar­ket rese­arch data in the nar­rower sense (pri­mary data from sur­veys) and insights from other data sources (social media moni­to­ring, panel data, online per­for­mance, etc.). In times of data-based value crea­tion, it seems cum­ber­some and inef­fi­ci­ent to keep dif­fe­rent data sources and thus also know­ledge reser­voirs sepa­rate from each other and available for ana­ly­ti­cal approa­ches. Howe­ver, pro­ces­ses and struc­tures that have deve­lo­ped over deca­des can­not be repla­ced by a sin­gle, defi­nable initia­tive.

Innovative approach required

What is nee­ded is an inno­va­tive approach that takes into account the spe­ci­fic cir­cum­s­tances of mar­ket and con­su­mer data. How can this be achie­ved, espe­ci­ally in medium-sized com­pa­nies with limi­ted resour­ces?

Linking added value in large companies

In large com­pa­nies, the acti­vi­ties and results of lin­king the added value that can be gene­ra­ted from data are evi­dent: for exam­ple, by poo­ling exper­tise and methods in data intel­li­gence cen­tres, where ana­lysts, data sci­en­tists and AI spe­cia­lists work tog­e­ther on tools and methods for the sys­te­ma­tic use of data in all areas of the com­pany. The Con­su­mer and Mar­ket Rese­arch depart­ment is also rapidly dif­fe­ren­tia­ting its­elf in this respect: Data sci­en­tists are beco­ming and are per­ma­nent mem­bers of teams, and the skills and fields of acti­vity of these depart­ments are evol­ving so that the dua­lity men­tio­ned above is wea­k­e­ning, data sets and know­ledge are con­ver­ging and value-adding pro­ces­ses are being stan­dar­di­sed.

Resource-intensive structural expansion

Estab­li­shing and expan­ding such struc­tures is resource-inten­sive:

  • The IT and data struc­ture must be coor­di­na­ted,
  • inter­nal and exter­nal experts must be brought tog­e­ther,
  • and data must be made acces­si­ble to users.

Responsible support and management

Anyone who wants to sup­port and manage such pro­ces­ses respon­si­bly is aware of the uncer­tain­ties asso­cia­ted with set­ting up a data lake or warehouse and imple­men­ting ana­ly­ti­cal solu­ti­ons, for exam­ple. Wit­hout a tar­ge­ted initia­tive, there is hardly any capa­city in day-to-day busi­ness to drive for­ward pro­jects of this kind strin­gently and with the neces­sary strength and resour­ces.

Market researchers as part of data-driven teams

Mar­ket rese­ar­chers are inher­ently part of data-dri­ven teams. They are respon­si­ble for and capa­ble of lea­ding these pro­jects and imple­men­ting change. They are ambassa­dors for data liter­acy and data demo­cra­tis­a­tion, i.e. they create access to data and ana­ly­ses for many people from dif­fe­rent areas of the com­pany. The aim of these pro­ces­ses is to gain know­ledge and, on the basis of data, to create value so that as many people as pos­si­ble with dif­fe­rent per­spec­ti­ves can con­tri­bute and uti­lise it in the best pos­si­ble and useful way for com­pa­nies. Hacka­thons are a good exam­ple: Giving access to people with exper­tise can lead to crea­tive and effi­ci­ent solu­ti­ons.

Sovereignty of companies over their data

At the same time, demo­cra­tis­a­tion requi­res com­pa­nies to have sove­reig­nty over their data. What may seem obvious can actually be a hurdle. This is because pro­vi­ders of exter­nal data are curr­ently try­ing to bind com­pa­nies to their own ana­ly­tics plat­forms (and thus keep them in a “clo­sed loop” or “lock-in”) by impo­sing rest­ric­tions or obs­ta­cles on the use of data. In col­la­bo­ra­tion with other inter­nal stake­hol­ders, such as cate­gory manage­ment, it makes sense to deve­lop a sus­tainable stra­tegy that pre­ser­ves inde­pen­dent use, i.e. data sove­reig­nty. How else can fle­xi­bi­lity be crea­ted that ensu­res adapt­a­tion to future requi­re­ments, pre­dic­ta­ble expen­dit­ure and sca­la­bi­lity?

Important aspects for evaluating and planning the data strategy

Break down silos: In many orga­ni­sa­ti­ons, data can­not yet be used effi­ci­ently because indi­vi­dual depart­ments own it or the qua­lity of the data makes it dif­fi­cult to access.

It the­r­e­fore can­not be empha­sised often enough that data har­bours a poten­tial that goes far bey­ond what the cur­rent (‘tra­di­tio­nal’) owners gene­rate in the com­pany. Col­la­bo­ra­tion bet­ween teams and dif­fe­rent data sources can often gene­rate signi­fi­cant added value.

A sin­gle point of truth: The crea­tion of a uni­form and bin­ding data­base ensu­res that all stake­hol­ders have access to the same data and that the ana­ly­ses and pro­ces­ses based on it, e.g. for report­ing, are trans­pa­rent and com­pre­hen­si­ble. This also includes auto­ma­ted data cle­an­sing and har­mo­ni­sa­tion (pre-pro­ces­sing). This increa­ses data qua­lity and work effi­ci­ency.

Auto­ma­tion of data pro­ces­ses: In order to gene­rate fin­dings with added value, pro­ces­ses should be digitalised/standardised where pos­si­ble (data­base queries, recur­ring report­ing). On the one hand, this gua­ran­tees the vali­dity of the ana­ly­ses. On the other hand, mar­ket rese­ar­chers can con­cen­trate on con­tent-rela­ted and advi­sory work.

Sca­la­bi­lity: Orga­ni­sa­ti­ons, mar­kets and tech­no­lo­gies are in a state of flux. Data appli­ca­ti­ons and stra­te­gies should the­r­e­fore always bear in mind that future requi­re­ments and needs can be met.

Mar­ket and con­su­mer ana­ly­sis teams can be given tar­ge­ted sup­port to set up new sys­tems. The focus is on empower­ment throug­hout: the exis­ting con­tent exper­tise is opti­mally com­ple­men­ted by the (auto­ma­ted) con­so­li­da­tion of various source data, stan­dar­di­sed pro­ces­ses and on-the-job ana­ly­tics trai­ning. As a result, the team emer­ges from the trans­for­ma­tion phase stron­ger, more crea­tive and more effi­ci­ent and can deve­lop its full poten­tial. Working with data is finally fun again.