The war is on: Real time versus predictive analytics

Banking customer analytics hierarchyBanking customer analytics hierarchy

The war is on... real time analytics using streaming data to responsively react to customer behavior with relevant and timely content versus curated data and predictive analytics.

At this stage the battles are increasingly being won by the real time side. Vendors like offer compelling abilities to take in huge amounts of data at near real time speeds and apply optimization algorithms that improve over time (as more data and outcomes are added, the models improve a.k.a machine learning).

The more traditional predictive and descriptive analytics that formed the backbone of CRM strategy for the last 15 years appear to be losing the battle, which translates into funding freezes and a lack of executive support and attention in many firms. However these capabilities still deliver the lion's share of growth drivers in terms of customer acquisition, cross selling and retention results.

It looks like a permanent transition underway - but I predict this will not hold true in the longer term.

    Real time is winning the battles but it isn't going to win the war.

One problem with real time analytics using an external vendor is your vendor will use your data - and all their other client's data - to optimize their algorithms. So your competitors gain the benefits of your data collection and management investment. To optimize real time analytics you need to share all your useful data, and for many Banks and Credit Unions this is akin to giving away the keys to the vault.

Another problem - and I think this is even more serious - is you lose control over your brand. If you are using the same decision engines and data as your competitors aren't you essentially cloning customer interactions across your industry? This may be fine for some commodity businesses, but in any industry where there truly exists a customer / brand interaction this sameness is a great way to kill a brand.

Having a deep understanding of customers and their aspirations, needs and behaviors over time is essential to creating a branded customer interaction and having a meaningful (as opposed to transactional) relationship with your customers. The only way to gain this level of insight is to have trusted, curated data, with a substantial time series dimension and using descriptive and predictive analytics techniques to truly understand and differentiate customer experience.

So I think we are currently in a bit of a hype cycle phenomenon which will drive the adoption of real time analytics beyond it's long term "share" of influence in customer interaction. That's okay to get it ramped up fast, but a longer term view suggests that investment in the more traditional analytics of data that you've already got in-house will continue to perform a meaningful role and lasting value to your business. Don't throw the baby out with the bathwater !

Questions? Please contact us