Predictive Analytics Works Well in Remarketing, So Why Not in Recalls?

Many auto­mo­tive exec­u­tives expect that more vehi­cle recalls will take place in 2015 and 2016. They may not be using the types of pre­dic­tive ana­lyt­ics tools that will be nec­es­sary for recall response and prod­uct plan­ning, accord­ing to a study by Deloitte. It may be a good oppor­tu­ni­ty for automak­ers to adopt sim­i­lar meth­ods to that which has become com­mon in used vehi­cle val­u­a­tion fore­casts.

In the Deloitte study, 42.3% of auto­mo­tive exec­u­tives who took an online poll expect more indus­try recalls in 2015 and 2016; but only 8% use advanced pre­dic­tive ana­lyt­ics to help pre­vent, pre­pare for, and man­age recalls. Near­ly one-quar­ter (23%) have no oper­a­tional prod­uct safe­ty and recall antic­i­pa­to­ry ana­lyt­ic capa­bil­i­ties.

“By cross-source cor­re­lat­ing inter­nal and exter­nal data sources, employ­ing spe­cial­ized advanced ana­lyt­ics, and lever­ag­ing inter­ac­tive visu­al­iza­tions, these com­pa­nies can improve cus­tomer sat­is­fac­tion, vehi­cle safe­ty, and brand per­cep­tions,” said Deloitte’s Derek Snaidauf.

We’ve seen used vehi­cle val­u­a­tion fore­casts change in near­ly all the mar­ket reports in the past two years. Recalls have been care­ful­ly fol­lowed on the remar­ket­ing side – as it very much affects brand qual­i­ty and new vehi­cle sales, and even­tu­al used vehi­cle val­ues. If vehi­cle recalls con­tin­ue, the pre­dic­tive ana­lyt­ics prac­tices will need to be inte­grat­ed more in recall response and new vehi­cle prod­uct plan­ning.

Near­ly 91% of respon­dents think that recalls are impact­ing the work­ing rela­tion­ships between sup­pli­ers and OEMs. One of the biggest chal­lenges to OEM-sup­pli­er col­lab­o­ra­tion in pre­vent­ing, prepar­ing for and man­ag­ing recall-relat­ed events was inef­fec­tive com­mu­ni­ca­tion chan­nels, accord­ing to near­ly 22% of the sur­vey respon­dents.




    Leave a Reply

    Your email address will not be published. Required field are marked *.