We performed exploratory data analysis to understand and profile customer and behavioral data and formulate the best analytic approach. We developed a machine learning model to dynamically predict the probability of voluntary (customer initiated) and involuntary (client initiated for non-payment) churn at an individual customer level. The outputs from the model also provided valuable insights into profile characteristics of churners vs. retainers.

The model enabled our client to make significant changes in acquisition tactics to shift marketing and spend from high churn channels and profiles resulting in higher early tenure retention and lower costs. The predictive model was used to focus retention efforts on at-risk customers and focus up sell programs on those found to be secure. These efficiencies allowed our client to target at-risk customers at a nearly 45% cost savings.

Tru-Clarity developed a segmentation model using customer profile, engagement and purchase data as well as 3rd party demographics. Five customer groups with distinct behavioral and profile characteristics were identified from the data. Key determinants of group membership were level of purchases, tenure, and engagement history including recency, frequency, channel and content.

From these insights, a lifecycle strategy was developed for onboarding of new customers, nurturing to encourage initial engagements across multiple channels, retention programs targeting at-risk points in the customer journey and personalized up sell of tenured customers to increase value.

We developed a Lifetime Value (LTV) model to forecast revenue, costs and profitability for an acquisition cohort over a ten year period. Retention curves at a sales channel, affiliate and demographic level were modeled from historic data. Fixed and variable costs were considered in the model so that a profit and loss statement (P&L) could be generated.

The client used this model to quickly assess acquisition costs versus quality. They terminated a number of unprofitable affiliate relationships and renegotiated terms on low to marginal affiliates. Marketing spend was optimally redeployed to the most profitable channels. The model has served a number of additional purposes including providing insights into cash flow and breakeven point by acquisition cohort, supporting revenue forecasting and analyzing variances of actual versus budget.

During the exploratory analysis phase of this project, we worked with the client to refine the definition of “best profile” as a combination of tenure and successful payment criteria. We developed a binary logit model using machine learning techniques to apply a score to a 3rd party data set indicating likelihood of matching the best customer profile. The model used demographic, behavioral and geographic attributes.

The client used insights from the model to support media buys to better reach their desired prospect profiles. The model scores were used for direct mail campaigns that efficiently reach their target audience. Using the model reduced direct marketing data spend by nearly 50%.

Using historic pricing and sales history, we developed a price elasticity model to estimate base and promotional price elasticity by product considering seasonality, day of week, sales channel, acquisition offer and other factors. Numerous internal metrics of existing customers were also analyzed including tenure, cyclical spending behaviors and total revenue. The model quantified expected change in revenue and volume for a given change in price.

The model strongly indicated that customer demand was inelastic on most products meaning the client could raise prices with the higher revenue from the price increase more than offsetting the decrease in sales volume. We also delivered a simulation tool to support sensitivity analysis around various scenarios. The client conducted extensive price testing and then rolled out across their product lines. They were able to increase pricing by nearly 15% in less than 9 months.

Tru-Clarity developed an analytic data set from all marketing activities including email campaigns, conferences, webinars and white-paper downloads. We performed a match back from the data set to sales opportunities and sales conversions and also appended industry code and company size to each record. From this process we identified specific campaigns that consistently yielded the best opportunity and conversion rates and highlighted many sources to scale back or eliminate. In addition to the industry and size data, we developed profiles for the best prospect targets with higher engagement and conversion rates and higher revenue levels. We also worked with the Salesforce administrator to develop better data management practices and training for the sales team.

As a result of this analysis, the client reallocated over $1 million in marketing spend to more productive use, implemented a more disciplined targeting strategy and revamped data management practices in Salesforce.

The findings in our initial assessment included insights into their customer base that were previously unattainable. The value of these findings provided our client with the leverage needed to push forward with key aspects of the BI roadmap. In assisting with these efforts, our consultants worked with the internal IT department to obtain access to structured data in a repeatable manner.

Initially, Tru-Clarity developed a data mart that resided external to the client’s data architecture. After several years we helped them create the data mart within their firewall with full internal IT support. Our role became that of user representative where we provided business requirements, user testing and project management services.

Tru-Clarity consultants start engagements such as this by learning as much as possible about the details unique to the client’s situation. We interviewed stakeholders and key users from multiple organizations as well as numerous members of their IT organization. We also hosted a Business Intelligence Workshop that allowed the stakeholders to share concerns. The client’s largest challenge was balancing the goal of creating a unified data source while segmenting PII data to protect citizen privacy.

With our guidance, the client adjusted their architecture to have a tiered approach that allowed for organizational customization with an underlying single source of citizen data. The operational data stores (ODS) and user facing data marts were unique to organizational requirements. The data mart maintained unified business rules and normalized information. Additionally, Tru-Clarity provided a best practice document that suggested a variety of ongoing changes including the establishment of a user community, improved quality control and the creation of a structured data governance program.

Tru-Clarity provided an experienced marketing consultant who was responsible for campaign reporting and functioned as an extension of the client’s marketing team. The consultant was responsible for generating daily, weekly and monthly campaign reports and providing commentary on noteworthy variances and trends. The consultant assisted in test design, evaluated test performance and made recommendations for new test and campaign design. The consultant prepared and presented a PowerPoint at each month’s operations review highlighting campaign activities and results.

The client was able to maintain continuity in their marketing department with this temporary consultant and benefited from enhanced reporting developed by the consultant and many insightful recommendations and analyses.

In the initial project planning and scoping discussions, we identified key challenges to be limited documentation of the legacy system, limited organizational understanding of the underlying data structures and content and regional variations in data capture. We developed an overall project plan with specific control points for the key challenges and risks. A thorough data discovery process was conducted to gain a comprehensive understanding of data content as well as data transformations and cleansing necessary to map the data to the target system. We conducted the production migration on a region-by-region basis to allow sufficient time for end-user testing and any necessary remediation. We implemented an overall governance process with key stakeholders and team members. Constant communication around timeline, status, deliverables, risks and hurdles was maintained. We held weekly project calls and interim checkpoints as needed and provided status updates and highlighted action items and risks. We also helped bridge some of the business and technical domain gaps among stakeholders.

The migration was completed on time and budget with minimal scope changes to accommodate additional data cleansing identified in the data discovery and testing phases. In addition, we helped the client set standards for data capture and business rules on the new platform and provided valuable recommendations for data quality management.