How to improve the effectiveness of your advertising

A major brand has carried out a large advertising campaign and has recently appointed Pointlogic to evaluate its effectiveness. During the campaign the advertiser undertook weekly interviews among its target audience to track the effectiveness of its campaign. Among other things, the advertiser asked the respondents about their awareness and consideration of different brands. The graph below shows how the brand awareness level evolved during the campaign, and how much media was used. The brand wanted to know how effective their campaign was and if they could have improved it even further using another media strategy.

 

Figure 1 shows the results on an aggregated level, but for the brand to really understand how their media strategy worked, we need to drill down to the respondent level data. Pointlogic have developed a method to untangle the data called Contact Estimator. For the respondents interviewed during the campaign, we can estimate how many exposures they have had with the different media used. This can be done by combining the stated media behaviour (‘I watch a lot of TV’) with the actual spend on the different media (e.g. 80 GRPs were spent on TV in week 2).

 

By combining these two pieces of information, you can get a first view on how the campaign exposures have impacted the effect metrics. The graph below shows the brand awareness level plotted against the number of media exposures.

 

This graph gives us some further insights into how media have worked. There is a clear positive relation between exposures and brand awareness: people who have seen more advertisements have a higher level of brand awareness. We can also see that this effect is neither linear nor proportional to the exposures: if we compare the effect at 4 exposures with the effect at 2 exposures, it’s clear that 4 exposures do give a higher awareness level, but it’s not twice as much as at 2 exposures. This phenomenon is called ‘diminishing returns’: for higher levels of exposures the effect is higher, but this effect diminishes as exposures increase. In practice this can be depicted by what is called a response curve.

 

When you split the respondents into two groups, people active in the category and people who have a passive interest in the category, you can also see that exposures among active people are more effective than among passive people. More exposures do not lead to significantly higher awareness levels among passive-interest people.

 

 

The graphs above show the effect based on total number of exposures, but the advertiser wanted to know the contribution of the individual media to brand awareness. By applying advanced statistical techniques we can derive response curves for each medium.

 

Figure 4 shows the response curves for two of the media used in this campaign (TV and Print) for people active in the category. The campaign also involved other channels but, for ease of illustration, we have only shown TV and print. The stars indicate the average number of exposures that respondents have had to this campaign. From this graph we can see that TV contributes most to the awareness level. We can also see that it’s not recommended to have high exposure levels, in this case not more than about 2. Exposures beyond the second one do not really add much to the awareness level. In this specific case, the brand could have improved the results of the campaign by using more TV and cutting down the Print budget.

 

In summary, this case study shows how using advanced statistical techniques has led to a better understanding of the effectiveness of the brand’s advertising campaign. The analysis produced a couple of valuable insights for the brand. Firstly, the campaign does appeal to people active in the category, leading to higher brand awareness levels, whereas the effect on people not involved in the category is limited. Secondly, the analysis shows that, in this case, a higher investment in Print has a limited effect and additional budget has a higher effect when invested in more TV exposures.

 

To find out more contact Erik ten Oever ( This e-mail address is being protected from spambots. You need JavaScript enabled to view it ).