In the world of digital marketing, where channels are constantly increasing, TV is still one of the main channels for some advertisers. Yet, it is still difficult to highlight its added value. How do you know if TV campaigns have a real impact on the consumer? And if certain anomalies of traffic on your site could find explanation precisely in the impact this type of campaigns?
Louis Janin, technical expert at Eulerian Technologies, brings you here some elements of answer.
Where to start ? What kind of data or information are we looking for?
Louis Janin : If a marketer wants to draw real conclusions from his analysis, focusing only on averages is the worst thing he could do. Looking beyond the masses is the key to find insights. For this, he must carefully observe any type of anomaly, whether positive (increased visits, carts, sales) or negative (traffic drop, clicks, etc.).
To draw conclusions, there is no need for a huge amount of data, but quality data, which tells interesting things. In the specific case of one of our retail customers (who had set up a national TV campaign), a single alert – for a drastic increase in hits, visits and carts – was enough to know that something was wrong on the site.
What to do when observing this type of “anomaly”?
LJ : When observing this type of anomaly in the data, we must investigate to know the cause. In the case of the retailer, the alert indicated an increase of 3,000% of visits, 1,500% of hits and 300% of carts. Faced with this kind of peaks, it is recommended to analyze the reports in the interface and verify that the data is correct.
Then we can go to hypotheses. This increase in activity on a website may, in absolute terms, be due to fraudulent traffic or even poor implementation of tracking. In this case, it was the impact of his TV campaign.
After this identification have been established, it is possible to explore several tracks to accurately analyze the performance of the campaign. In our example, the most relevant information was the total audience that could see the spot (3 million viewers) and the original channels of traffic to the website during its broadcast (Direct Access, SEA and SEO ).
What to do with the results of this analysis?
LJ : The results will serve different purposes: to prove that investments in some channels (rather than other) were justified and to continue to invest in them, or to favor channels that have performed better than expected. The notion of duration of impact is also interesting. If we take our example, for the next TV campaign, the retailer can:
- schedule a dedicated campaign related to the spot just the time of the impact (2 hours), or outbid some of its AdWords campaigns (since the SEA is a channel that is highly impacted)
- or on the contrary, cut SEA campaigns for 2 hours if it is estimated that the users click on the SEA instead of clicking on the SEO, but that they would come on the website anyway
Moreover, to avoid considering an outperformance of the channels impacted over the period, the “broadcast” channel can be activated. This one will absorb the visits considered incremental compared to the norm (defined upstream by the advertiser or calculated as average of visits on similar days).
In any case, analyzing the data will provide a clear understanding of what works and what does not work in the marketplace. A marketer, who can review data, has the potential to evolve a traditional business into a data-driven business.
To learn more about the analysis of TV campaigns and their impact on an e-commerce site, you can consult our Use Case Promofarma here.