A leading car manufacturer.
The client's fleet management division contributed approximately 33% to its overall revenues with potentially untapped segments. However, the growing adoption rate of alternate fuel vehicles had resulted in a continuous decline in profitability.
The client decided to respond to the challenge by empowering its fleet management division with advanced digital and consumer analytics tools to attract new customers as well as to retain existing customers.
The client had already identified that corporate and leasing agencies were high-value customers that needed encouragement and focus to minimize any churn since it could lead to a major drop in revenues. SG Analytics helped the client come up with business requirements and analysis to narrow down their issues, to the following solutions:
Firmographic analysis which was used to understand growth opportunities for corporate fleet and offers.
Integrated market-research-based approach was used for identifying potential customer targets.
Lead prioritization methodology based on the sales funnel and marketing activities were prioritized.
SG Analytics' team helped the client implement a campaign management with micro-segmentation and targeting using the IBM product suite.
The team also enabled a better customer focus by allowing to target each segment separately through internal and external databases.
We utilized predictive analytics to build the following models:
X-sell/up-sell models based on the next best product propensity with market basket analysis using an Apriori algorithm. The final model was chosen based on the confusion matrix and model lift generated.
Campaign response modeling based on logistic regression, utilizing carefully chosen variables including CLTV and churn probability of the final segment.
Survival modeling for high-value customers using hazard function calculating the survival probability on a monthly basis and help predict churn over a period of time.
Improved acquisition efficiency for direct mail by 4x.
Increased email efficiency by improving cadence/suppression.
Learning algorithm runs daily to ingest new data and deliver scoring to existing systems.