![]() | | | | | |  |
The Science of
Customer Experience Transformation
RESULTS
Since its launch Insight has proven to be more effective and precise in its predictive performance than any analytic technology or approach thus far encountered. Effectively Insight is capable of detecting patterns in data that current analytic technologies miss.
In every instance when Insight’s results have been compared to results produced from traditional approaches Insight has performed better when measured from any perspective. These results have been:
1. Profitability (product, portfolio, category)
2. Acquisition (sales, revenue, market share)
3. Media channel (profiles, formats)
4. Segment (preference, pathways)
5. Campaign (combinations, sequences)
6. Investment (budget, forecast, optimal)
Insight’s performance has been demonstrated in the results obtained by using Insight compared to existing methods. Having performed comparative tests many times, a definitive empirical test has become an essential aspect on any decision regarding the use of Insight.
|
| Returning a Company to Profitability |
Canada’s second largest insurer operated several companies under their banner. Within the group a number of companies were not performing profitably. Furthermore in Canada regulations were emerging to prevent the use of credit scores as a base for ranking customers making it difficult to determine who to insure and at what price.
The Engagement
Using only the same data as the insurer’s analytics team within a week Insight produced a set of customer scores that more accurately and reliably assessed the risk profile of each customer. The insurer assessed that there was a greater than 20% improvement in accuracy in customer profiling over the previous results. Insight then produced campaign and event trigger executable rules to apply to customers using the insurer’s CRM system. The insurer’s analysts estimated a $40 million annual improvement. The insurer returned to profitability.
|
Removing the Risk of a Take-Over |
One of America’s largest mid-western insurers was in danger of a takeover or sale due to severe profitability deterioration over a number of years in homeowners insurance. The team was engaged to develop both customer segmentation and pricing models to replace the actuary developed segment and pricing models.
The Engagement
Using the existing database within a week Insight analysed the last three years customer data to score the insurer’s entire customer portfolio. The scores were then banded into 13 distinct groups ranked from the most profitable to the most loss making. Insight’s scores helped identify the best offers for each product. Very simple changes were then made to the insurer’s pricing system in order to correctly price customers. Insight provided a very significant improvement over the insurer’s actuarially derived pricing and addressed the adverse customer offer strategy and returned the insurer to profitability in 18 months while preserving market share.
|
Managing the Impact of a Completely New Industry Scenario |
The leading North Eastern US Insurer was faced with significant business instability due to regulatory changes impacting the entire industry. These changes would render the insurers’ current analytic methods ineffective due to limited relevant data being available on how regulations would impact both competitors and consumers.
Using the insurer’s existing database Insight produced a set of customer scores incorporating very limited predictive data. The scores more accurately assessed the risk profile of each customer as a result of the regulatory change. Insight’s scores helped identify the best customer offers for each product. Adjustments were made to the insurer’s pricing system to correctly price customers. On the change-over date to the new regulations the insurer had a significant jump on the other “big 10” competitors in this state. At a time of regulatory instability the insurer remained market leader in the state and actually increased market-share
|
|
Managing the Risk of Fraud
|
A large Australian bank supplied details of credit card customers of which approximately 1.6% had defrauded or defaulted in the first 12-18 months. Automated statistical evaluation has become the mainstream means of evaluating both willingness and capacity to repay debt. Known as ‘scoring’, these statistically derived models used to assess credit card applications are based on data available to predict a customer’s credit performance, and have proven to be more effective, and to remove the subjective nature of assessments. Running costs associated with building, maintaining and operating scoring systems are estimated to exceed $10 million annually for the major banks.
Using only the same data as the bank’s analysis within a week Insight produced a more accurate model that identified the most likely 10% to defraud and identified 70% of all card defaults. The model represented a greater than 10% improvement over the bank’s internal results developed by its expert teams. Insight produced customer treatment rules that were executed across the banks internal systems. Independent analysis by external consultants predicted a $335 million improvement in the Bank’s bottom-line over 5 years.
|
|
|
|
|
| | | | |
| ![]() |
