19 January 2018
Paul Russell

75846

Paul Russell - Experian

3Posts 9,111Views 0Comments

Is there a place for social media data in credit risk assessments?

16 May 2016  |  4397 views  |  1

When I ask a lender for credit, there are two questions the lender would like answered: is Paul willing to pay and is he able to pay? Most lenders will try to answer these questions by taking data about me and applying some techniques from mathematics to my data.  They do this to estimate the likelihood of me not paying (a technique often called credit scoring) and to estimate how much disposable income I have (often called an affordability assessment). While the mathematicians have made advances over the years in estimating willingness and ability to pay, it is the availability of more data about all of us that has driven the improvements in lenders’ ability to decide who is both willing and able to repay their credit.  

We all go through our lives leaving a trail of data in our wake.  Some commentators have called this ‘data exhaust’. This has always been the case, but in advances in technology and our appetite to use it mean that more of our everyday behaviour is captured by computers and we are all giving off more exhaust than we used to.  Our use of social media is a good example of new behaviours generating new data.  

But it’s not all about social media.  Many of the firms we deal with are generating their own data – recordings of customer service calls, visits to a firm’s website, preferences expressed through these sites, delivery histories…etc.  It all says something about us as individuals.  But of course, it needs to be relevant.  

Social media data could fill holes in the data available to credit scoring particularly where there is little other information about a person eg young people who have yet to build up a credit history.  It may be of more use in countries where robust credit history based systems do not exist or there is no mature credit bureau. 

It is worth noting, however, that social media data is not widely used for the assessment of credit risk at the moment and is not used at all by us.  While some of these new data sources (whether they’re external to or internal to an organisation) might be useful in predicting credit risk, we feel that the impact will be marginal. 

We should not forget that there are rules governing how data can be used in credit decisions.  When considering new sources of data we need to answer a number of questions: has the person that the data belongs to given their consent for the data to be used?  Who owns the data and can the data be used for the purposes of credit scoring?

Consideration should always be given to what data should be used and for what purpose. 

For the majority of UK adults, the predictive power of credit account data, supplemented with bank account transaction data remains the strongest basis for credit decisions.  They provide lenders with high quality, consistently structured data with clear and reliable provenance.  In simple terms, data about what people do (or have done) when using credit is currently a more powerful predictor of future credit behaviour than what they say they do (or have done) in other parts of their lives.

Nonetheless, major organisations across all industries have taken data seriously for a long time. Many are using social media and website visits to understand what people are interested in and how they interact with organisations. Organisations are always looking for new data to help them better understand people, not just when it comes to credit decisions.

TagsRisk & regulationRetail banking

Comments: (1)

Konstantin Rabin
Konstantin Rabin - Kontomatik - Warsaw | 16 May, 2016, 11:42

I definitely agree with "Social media data could fill holes in the data available to credit scoring particularly where there is little other information about a person eg young people who have yet to build up a credit history".

On its own, Social media data is unlikely to be the only source of information for credit asssessment. There are a few reasons for it, but I feel that the most important one is that it is pretty easy to manipulate the data.

However, if social media data is combined with another powerful source, like a banking data of the user, it can certainly be a very useful.

By the way, there are companies that have been using social media data for scoring for a few years already :)

Would be lovely if you could provide a few exmaples of using the data though. Apart from that, great job!

Be the first to give this comment the thumbs up 0 thumb ups! (Log in to thumb up)
Comment on this story (membership required)

Latest posts from Paul

Is there a place for social media data in credit risk assessments?

16 May 2016  |  4397 views  |  1 comments | recomends Recommends 0 TagsRisk & regulationRetail banking

Reaching those tech savvy customers

28 July 2015  |  2307 views  |  0 comments | recomends Recommends 0 TagsMobile & onlineInnovation

What’s in a wristband?

24 June 2015  |  2408 views  |  1 comments | recomends Recommends 0 TagsPaymentsInnovation

Paul's profile

job title Director of Analytical Solutions
location Nottingham
member since 2015
Summary profile See full profile »

Paul's expertise

Member since 2015
3 posts0 comments
What Paul reads
Paul's blog archive
2016 (1)2015 (2)

Who's commenting on Paul's posts