The relatively high costs of television advertising mean that businesses should take extra care that every pound is well spent. StatCore worked with a major high street bank to test the effectiveness of their television adverts and to ensure that there was a predictive relationship between the adverts’ screening and sales increases. StatCore’s challenge was to:
- Consult with a variety of key stakeholders in the business to find out all the different factors that could have been predictive of sales; over 100 variables were tested, e.g. changes to the product, other product launches, price changes.
- Research and incorporate external data on other variables that could have led to a difference in sales, e.g. actions by competitors, seasonality, the advert’s decay rate (how long its stays in the customer’s mind).
- Build econometric models for a range of products to estimate how many sales resulted from the television adverts. We do this by:
- comparing how a range of variables behaved in the past
- correlating those behaviours with weekly sales both before the advert and after.
- Quantify halo effects of advertising – where the advertising of one product (e.g. credit cards) positively affected the sales of another (e.g. loans)
- Communicate results to non-technical audiences, using charts to visually represent the time series models.
Many businesses outsource this advanced level of analysis, however working with the client in-house over nine months meant that StatCore could:
- Produce richer, more sophisticated results through daily consultation with stakeholders.
- Devolve knowledge and skills to the in-house team
- Offer a better value service
The bank was able to accurately calculate their Return On Investment for the television campaign and make a confident decision to continue running similar television advertising. StatCore’s work also informed a company -wide optimisation project to determine which of their products should take centre stage in television advertising for the future.