Big Data – Big Problems?

Lego Big Data AnalyticsI was asked an interesting question this week about whether the recent controversy around governments gathering data to map and inform security services would impact the current themes of social media, big data and analytics in the world of business.

My answer was ‘Yes’ and ‘No’.

Leaving any wider politics around this issue aside, this has become a burgeoning industry in recent years. Our ability to store data has increased dramatically and we have been finding ways to use this data to add value to business.  We are all consuming ever more data.

Is this new? Certainly not business has always run off data.  However a public spotlight has been sharply focused on the technique. This will undoubtedly raise concerns; concerns that need to be addressed for the public to be made comfortable and negate any impacts.

So here are a couple of good questions to ask

Does it add value to the customer? Good examples here are Amazon and Netflix movie recommendations. I like these recommendations, they are based on my and other customers purchases. This is data-analytics in action.  It genuinely adds value to me, my experience and therefore I am comfortable.

Can customers understand the linkage? Are the linkages between observations and outcomes logical, something that is intuitive that I understand. Ie ‘Because you bought product x, you may like product y too’ or more close to home for me ‘because you have been late on payments before, we anticipate you maybe late in the future’. These are all logical, I may not like the answer, but the logic is intuitive and therefore I can agree with.

What is the exception process? Any statistical model is not perfect, it cannot predict the future with 100% accuracy, so there needs to be process to gather new information and make updated decisions. Just because I was recommended this product based off the model, it doesn’t preclude me from purchasing something else.

There is a careful line to be drawn here for any business. Going too far, not being logical or being unyielding will be seen as intrusive and raise concerns in the customer base. Getting this right can add value to the customer, their loyalty and business profitability.

The bottom line here is customers will not mind, I believe, if we are responsible, using the data proportionately and for their best interest. In many ways just like shop keepers of old, the personal touch can help, but don’t be ‘creepy’.

So are there Big Problems for Big Data? I think the term Big Data will change (it is too scary for people), and there will be heightened awareness of the issue (cloud apps may struggle), but as long as we remain responsible, the approach will continue. The genie is out of the bottle and it adds too much value for all to stop.

Photo Credit: Dunechaser via Compfight cc

Posted in Opinion | Tagged , | Comments Off on Big Data – Big Problems?

CSRSA meeting

Lucky enough to be up at the Credit Scoring and Risk Strategy Association meeting this year.  A great event, location, and group of people.

A couple of key thoughts and take-aways from the sessions.

For all the recent talk of big data and analytics in the media, this industry has been doing this for years and is ahead on the curve.  Consequently much of the discussion was less about the ‘how to’ do this, more around ‘can we get more, new and interesting data’, that is predictive.  It is always refreshing.

Cost and performance continued to be big themes, with processed focused analytics (drilling to the next level, as a result of more data availability) and network analysis being the ‘newer’ themes discussed.  My feeling is the network analysis theme is one we are going to see develop and expand more as a discipline over the next couple of years.  (it can be very powerful and adaptive)

The last take away was around survey’s and industry comparison, this is always valuable, and really got the group engaged.

All in all very thought provoking and it was great to be there.

Photo Credit: Matthew Ingram via Flicker

Posted in Opinion | Tagged , | Comments Off on CSRSA meeting

Social media/extended data & Credit and Collections

Last week I was lucky enough to attend the annual CAGT conference in Toronto and speak on social media – data, with respect to the credit granting industry.

There was some great discussion around the economy, outsourced managed services and social media.

Economically we still seem to be somewhat in ‘wait and see’ mode, with slowing growth in new loans and corresponding delinquencies at all time lows (albeit with increased total debt).  In this environment there is still continued focus on cost and delivering performance with still a clear need for new ideas to continue to drive results.

Social media, is obviously one of those hot topic areas and discussion focused on a couple of themes: privacy and the need for consent together with some of the latest thinking on how to use and leverage social media to drive business and sales (Ie how to really use social media to drive value).

For me it feels while there has been much thinking and investment up stream in the sales and marketing space we are only now starting to see some of these ideas filtering through to credit granting and collections.  There is potential to provide improved performance from this data with a few innovative companies are leading this change.

As we gradually learn however, it is also clear there will need to be an important paradigm shift in the way we think about and relate to data.  For example;

At the moment the data is generally a consequence of the process.  We gather all the data  from each process and work out how this could be useful.  It is process centric data.

In the future state we will need to start thinking much more in terms of data centric processes: designing the data requirements into each process, thinking about the data gathered, its applicability, predictability, use and of course consent at every stage in the customer life-cycle.

This is of course easy to say and understand.  It reality it is often hard to do.

This will undoubtedly also drive increased transparency of actions across processes; linkages will need to be mapped, impacts understood.

It will however be critical.  Gathering this information and consent once a customer has run into an issue is usually difficult or impossible, especially if there is a loss of trust.  Gathering the information upstream could help in addressing the issue, even avoiding it all together.

The change will be tough, however done right has potential to transform the customer-business relationship for the better….

Posted in Opinion | Tagged , , | Comments Off on Social media/extended data & Credit and Collections