The big-data challenges facing financial services

Many finan­cial ser­vices firms have gained access to huge vol­umes of new data as a result of their dig­i­tal trans­for­ma­tions, but they’ve bare­ly start­ed to extract full val­ue from the wealth of mate­r­i­al at their dis­pos­al.

The aver­age com­pa­ny col­lects and analy­ses only 24% of the oper­a­tional data avail­able to it, accord­ing to a glob­al sur­vey by Sea­gate and the Inter­na­tion­al Data Cor­po­ra­tion in 2020. There are sev­er­al rea­sons for this, rang­ing from con­cerns about cus­tomer pri­va­cy and reg­u­la­to­ry com­pli­ance to dif­fi­cul­ties get­ting all the rel­e­vant mate­r­i­al in the right for­mat for crunch­ing. 

The first big chal­lenge fac­ing finan­cial ser­vices firms seek­ing to make more and bet­ter use of their data is “the data ecosys­tem, the sec­ond is tal­ent and the third is data man­age­ment”. That’s the view of Edouard Legrand, chief dig­i­tal offi­cer at BNP Paribas Asset Man­age­ment. 

The elas­tic, almost unlim­it­ed capac­i­ty of the cloud is help­ing to address the first chal­lenge. Cloud-based tech­nol­o­gy has also intro­duced many new capa­bil­i­ties to data and ana­lyt­ics teams. But Legrand believes that, “even if progress has been made regard­ing the extrac­tion, han­dling and gov­er­nance of data, there is always room for improve­ment. Most of the focus has shift­ed to imple­ment­ing cross-func­tion­al plat­forms that enable every­one in the organ­i­sa­tion to make the most of the data. Ulti­mate­ly, all infor­ma­tion should be avail­able seam­less­ly, as should the tools to exploit it.” 

When Legrand talks of the tal­ent chal­lenge, he’s refer­ring to the industry’s inabil­i­ty to attract enough peo­ple with the req­ui­site IT skills.

Attracting talent

Nick Broughton, chief infor­ma­tion offi­cer at Novu­na, agrees with this assess­ment. “Tech­nol­o­gy alone doesn’t gen­er­ate val­ue; you need data-lit­er­ate peo­ple with good ideas,” he says. “Data sci­ence skills in par­tic­u­lar are key to obtain­ing real insights from the wealth of data we have. Attract­ing, retain­ing and grow­ing our inter­nal tal­ent pool around these new skills is an addi­tion­al chal­lenge when the demand for them in our mar­ket is so high.” 

A sur­vey of more than 250 finan­cial ser­vices firms in Novem­ber 2021 by recruit­ment giant Hays revealed that 83% had strug­gled to recruit data sci­en­tists, even though they were typ­i­cal­ly offer­ing annu­al salaries of more than £100,000 for such jobs. More than a quar­ter of respon­dents report­ed that they didn’t have all the skills they required to achieve their com­mer­cial objec­tives. 

The sec­tor will have to become more flex­i­ble with its employ­ment poli­cies and prac­tices if it wants to attract and retain the data spe­cial­ists it so sore­ly needs. Know­ing that they are in such great demand, these pro­fes­sion­als can dic­tate the terms. Many pre­fer to work at home and not on the usu­al nine-to-five sched­ule, for instance, so it’s up to employ­ers to make allowances for that. 

Some firms are going to great lengths to estab­lish a reli­able pipeline of tal­ent by, for instance, estab­lish­ing rela­tion­ships with data sci­ence com­mu­ni­ties and set­ting up appren­tice­ship pro­grammes.

When it comes to tack­ling the third chal­lenge that Legrand cites – data man­age­ment – com­pa­nies are work­ing hard to put more effec­tive gov­er­nance sys­tems in place. Their over­all aim is to pro­vide a holis­tic view of their prod­ucts and cus­tomers using data col­lect­ed and processed in real time. It’s vital to make analy­sis tools use­able and eas­i­ly acces­si­ble to every­one in the organ­i­sa­tion who needs them, since no one wants to be call­ing the IT team for help every five min­utes. 

One of the main oppor­tu­ni­ties aris­ing from all this work is that it helps com­pa­nies to come up with new ways of sat­is­fy­ing their clients, accord­ing to Broughton. 

“Cre­at­ing out­stand­ing cus­tomer expe­ri­ences should be cen­tral to any data-dri­ven ini­tia­tive,” he says. 

The role of AI

Pro­vid­ing ever-more per­son­alised ser­vices is one obvi­ous area of devel­op­ment, but the poten­tial of aug­ment­ed advice and ser­vices, where­by data insights com­ple­ment human inter­ac­tions, is also excit­ing some com­pa­nies. By com­bin­ing sev­er­al sources of data, for instance, they can gen­er­ate invest­ment sig­nals and intel­li­gence that rela­tion­ship man­agers can use to keep their clients bet­ter informed. 

With the help of AI tech such as machine-learn­ing sys­tems, client-fac­ing employ­ees can iden­ti­fy pat­terns and trends that they would nev­er be able to spot unaid­ed. Intel­li­gent tools can also help cus­tomers to bet­ter under­stand their finan­cial health and the risk/return pro­files of the invest­ment oppor­tu­ni­ties avail­able to them. 

“We see a future where clients can use tools to exper­i­ment. AI mod­els can show them how invest­ments could change over time, for instance,” says James Brake, inter­im chief data offi­cer at Har­g­reaves Lans­down. “Yet we find our­selves in a world where reg­u­la­tion often lags behind tech­no­log­i­cal inno­va­tion. With this in mind, finan­cial ser­vices busi­ness­es must be care­ful in their approach to AI-enabled ser­vices, say, to pre­serve their clients’ trust. It’s why Har­g­reaves Lans­down recent­ly cre­at­ed the role of head of data ethics. This helps to ensure that our approach to AI is trans­par­ent, repeat­able, unbi­ased and able to deliv­er the best out­comes for clients.”

Finan­cial ser­vices firms are already using vir­tu­al assis­tants and chat­bots pow­ered by nat­ur­al lan­guage pro­cess­ing, an AI tech­nol­o­gy that’s fast becom­ing stan­dard fare. Nat­ur­al lan­guage pro­cess­ing also enables them to run auto­mat­ed search­es of infor­ma­tion sources rang­ing from news feeds to earn­ings reports, which will quick­ly iden­ti­fy poten­tial prob­lems such as prof­it warn­ings and instances of green­wash­ing. 

The use of alter­na­tive data sources – for instance, satel­lite cam­eras and smart sen­sors for cli­mate-sen­si­tive invest­ments – is also like­ly to become a stan­dard way to inform per­for­mance fore­casts, for instance. 

Open bank­ing with data-shar­ing is anoth­er excit­ing area to watch.   

“As the world moves towards think­ing about data as a prod­uct, we need to start build­ing our ser­vices with the data they serve in mind,” Broughton says. “Mak­ing our data prod­ucts more inter­op­er­a­ble, secure and gov­ern­able will be cru­cial to the suc­cess of future ini­tia­tives in this field.” 

Once data becomes the prod­uct, expect to see a whole new ball­game.