client:First Capital

Newspaper advertisements across the country resulted in calls each month to First Capital’s call centre. Due to the diversity of publications, placements within the paper and advertising day, it was difficult to gauge which ads and publications were the most successful at generating calls.

We created a detailed linear regression model on historical data that successfully measured the benefits of each publication and day of the week. The model also then forecast very accurately the number of incoming calls that would be generated by various advertising scenarios.

First Capital were then able to refine their advertising spend to ensure maximum return for their dollar.

Simple Example
In the table on the left, you can see that various ads were placed on different days of the week in each month. Within this, the ads were also placed in different publications around the country. Total calls coming into the call centre during each week and month were recorded but not necessarily allocated to any particular day.

It is quite easy, using sound and reliable mathematical techniques, to "work backwards" (using regression) to determine which combinations of days and publications had the greatest impact on total call volume.

The example results show that there were 4 days that had a strong impact and 3 that were less successful. The company could then refine their ad placements to concentrate on these days and get stronger ROI for their budget.