Simply connecting to data sources such as SAS, SPSS, MATLAB and R is only one piece of the puzzle.
- Completely streamlining a process to deliver measurable, timely, credible and meaningful insights closer to the customer’s decision is the vision.
- Managing the risk/reward equation to achieve sustainable competitive advantage, predictability and higher profitability is the goal.
Earlier in my career, I worked as a Pricing Manager in the auto finance industry. During my first weeks, the common topic of “Adverse Selection” and its negative impact on profitability, pricing response models and credit risk models was discussed in several meetings. This was a new term for me so I walked into my manager’s office and asked him to explain adverse selection to me. His answer:
“I’m not exactly sure how to explain it, but I know it when I see it”
He went on to describe the impact of pricing on the customer’s decision to buy. The biggest fear is that higher risk credits take advantage of poor pricing models and thus, blow up the ability to forecast losses properly. My passion for bringing complex analysis to the entire organization and deliver on the vision and the goal was born.
Bringing the Data Scientist Work to the Point of Decision:
For starters, building applications from data/output of traditional analytics software (SAS, SPSS, Matlab and R) is an extremely common use case in Financial Services. However, the analytics market often gets distracted/confused by the differences between "integration" and “connectivity". My response is yes, Qlik customers do both, plus a whole lot more.
Plus a whole lot more:
It is extremely common to connect directly to these data sources through available ODBC drivers or, export the model results of these calc engines to a file or warehouse. The Deloitte solution called “StingRay” is a great example.
The analytics engine runs scenarios, stores them in a data warehouse and is brought into QlikView either on-demand or on a set schedule. The janitorial spreadsheet (and spreadsheet add-on) work was eliminated from the process and allows the customer to run a scenario and obtain immediate feedback.
We also have customers who integrate with these tools by sending and receiving commands/routines from the Qlik user interface. This option has been used by several large global financial services firms and has the advantage of immediate feedback based on the user selections. Here is a link to a Qlik Community post that provides a deeper dive into this solution.
- Qlik’s powerful analytics engine has the capacity to take some or all of the workload on itself. Our customers have the option to combine the speed and power of running calculations in-memory or within the extract & transform stage, both are fast.
- Many financial services examples of utilizing these options go far beyond simply connecting and integrating: they can be found on our Industry website. For example, Qlik hosted a webinar with Deloitte where a full forecasting/valuation model was created inside of QlikView running Monte Carlo simulations on many variables and providing sensitivity and scenario analysis to a multiple year forecast.
Value is optimized though, when these model outputs are combined with other data sources in an associative model and provided to a broad population of business users.
Business users may not understand statistics but do understand that when an application tells them they are 'out of compliance,' 'under-utilized,' 'missing targets' or 'inefficient'.
Banks, insurance companies and security & investment firms are executing on the vision and achieving their goals through a governed, scalable, enterprise-class data discovery platform that goes far beyond tactically connecting to a single data source.
Photo credit: liza31337 / Foter / CC BY