The Roots of Predictability Are in a New Generation of Marketing Metrics

Two large financial institutions came to us with a very similar problem in measuring marketing. Their marketing organizations were under tremendous pressure to improve their financial accountability to their respective organizations. In each case, the immediate challenge was to provide a more holistic understanding of marketing performance rather than continuing to fixate on an elusive ROI calculation (a number that we find difficult to obtain and, sometimes, misleading).

The underlying objectives for both organizations were not only to measure marketing’s impact on sales and profitability but also, and more importantly, to …

• Determine the contribution of marketing to the overall firm
• Forecast and predict the planned impact of future initiatives

Because neither institution had a fully understood diagnosis and framework to predict the impact of marketing activities to financial, customer and marketing results, we bought the marketing executives to the table and essentially asked them, “If you had another million dollars to spend, what could you do with it and how confidently can you predict the results of the spend?” We find that this usually results in a one-time econometric ad hoc modeling exercise focused on forecasting the potential return on that incremental spend, out of context with an analysis of the underlying causes driving returns in marketing.

To answer our question, we worked with both organizations to systematically craft a metrics framework that aligned with corporate strategy and provided meaningful data to key stakeholders. That, in turn, has led to more actionable information and answers to incremental spend returns. The basis of this new generation of measurement framework was that it would provide the first predictability in marketing.

You achieve that first predictability through understanding the cause and effect in marketing. You tie your marketing activities (cause) with the marketing performance, financial results and customer impact (effect), thereby creating credibility within the discipline of marketing.

The new generation of marketing metrics framework enables organizations to measure marketing performance holistically and to increase predictability of marketing performance. This is absolutely a key part of any marketing executive’s toolkit.

In the case of the first financial institution, which we’ll call ABC Financial, the key objectives were to measure the marketing spend on acquisition, growing customer share of wallet and investable assets with the organization. To measure the marketing performance with metrics like “new accounts added” and “new households added,” we had ABC Financial marketing executives use direct measurement techniques that included doing cell-level measurement with statistical significance testing. To get more sophisticated metrics around the financial impact, such as “average increase in investable assets” and “growth in customer share of wallet,” we had them employ econometric modeling with statistical weighting techniques to derive the metrics from the underling marketing activity data with cell-level treatments and statistical significance testing layered on top of the marketing activity data.

Based on these results, ABC Financial marketing executives were able to segment their customer population into six different segments, with a clear migration pattern and the most profitable segment being the households with the highest investable assets. ABC Financial was able to look at its segments and was able to profile prospects and acquired customers to these segments. This ability to learn and segment the customers enabled ABC Financial to devise marketing strategies to progress customers from the lowest segment to the most profitable segments.

ABC Financial measured, forecasted and reliably increased its return on marketing investment by 18 percent after the first year of execution. By spending less money on fewer, but more effective, marketing activities, the company was able to optimize customer profitability and lifetime value through highly targeted cross-sell and retention programs.

Through a series of experiments, executives have a much deeper understanding of the optimal sequencing of marketing activities and best offer mix throughout the buy cycle, thereby maximizing the cumulative effect of their marketing efforts against a targeted customer segment. This insight has also increased their share of wallet with the most profitable customer segments by nearly 15 percent. They have also effectively analyzed and achieved deep insight into the type of customer acquired and the loyalty of the customers to the brand. This insight, which they were able to compare with their tenured customers and the loyalty of that segment, has helped them target the right kind of customers for acquisition. This initiative has led to increased overall increase in their product penetration by around 8 percent.

The objectives for the second financial institution, which we’ll call XYZ Financial, were to look at all three effects of the marketing activities in the above framework, with an emphasis on these factors:

• Direct measurement. This dimension correlates the marketing performance as a direct result (effect) of the marketing activities (cause). The measurements in this dimension included such metrics as “marketing spend on acquisition,” “spend on growth” and “spend on retention or attrition risk households.”
• Econometric modeling. This dimension analyzes the effect on financial results with respect to the marketing activities. The key metrics analyzed in this phase focused in revenue and profitability from newly acquired customers; revenue and profit growth from existing customers; projected saved revenue; and profit from at risk customers.
• Customer Touch-Point Interactions. This dimension focuses on analyzing the effect of marketing activities on customer impact. The customer impact metrics analyzed in this dimension included “customer lifetime value,” “percent of wallet share” and “customer loyalty metric for tenured and new customers.”

XYZ Financial was able to understand the financial impact of acquired households, saved households and financial impact of customer loyalty. Based on the results, XYZ Financial is going to alter its marketing spend significantly with focus on marketing programs on growing revenue and profit from existing customers and setting goals for improving the reduction in attrition rate by at least 10 percent.

There is no conceivable way that a company can build and scale this framework without, at least, basic resource management, campaign management and dashboard tools. The good news is that a lot of companies already own these applications. The bad news is that few have yet to figure out how to configure and tune them to achieve reliability, let alone predictability. And few dashboards inform reality or are truly predictive.

Dashboards, when applied with this framework, will be predictive in nature. The key takeaways for both of the financial companies we worked with are that they can achieve predictability by finding the drivers of cause-and-effect relationships in metrics and that the underlying analytics and dashboards provide a visualization of both the effects and root cause metrics that are the key drivers of performance aberrations.