How To

How TV campaign tracking works


This article in: Deutsch | Français | Italiano | Türkçe

TV-Analytics-wywy-website-visitsFor the first time, the so-called Second Screen users allow to track the impact of TV commercials in real-time as (a part of) these users instantly hit the advertisers’ websites seconds after the airing. In fact, based on our research about 80% of the hits will happen within the first 90 seconds.

This effect can be measured with an online analytics tool.
In order to start your TV campaign tracking, you will need two things:

1. An analytics tool to properly attribute the TV traffic to your website KPIs. Many analytics tools already offer an out-of-the-box TV advertising module. You can even do it with Google Analytics or with Adobe Analytics if you are willing to invest some time for the analysis.

2. Accurate TV airing data. You cannot rely on your TV media plan as the planned airing times are almost always a couple of minutes off. Depending on the analytics tool you can either use real-time data or ex-post csv files with exact TV airing times.
TV campaign tracking infographic

When the TV spot is aired, the analytics tool in a first step “flags” all incoming users with an additional tag “TV ad” for a certain time frame. This attribution time usually varies between 3 to 7 minutes. Our research has shown that on average >95% of visits due to the TV spot come within 5 minutes of the airing.

In a second step, the statistical analysis of the analytics tool then “separates” the regular users from the users who visited your website due to seeing the TV ad: The overall traffic will be subtracted from the baseline traffic. The remaining delta is then contributed as the uplift thanks to the TV ad. The baseline represents the “usual” traffic, e.g. by looking at the X minutes attribution time PRIOR to the spot airing. It is important to note that there is no attribution on an individual user level (as there is no way of telling for sure if a user watched TV) but on an aggregate level. The statistical analysis does the trick of “separating” users by calculating the delta between “normal” traffic and traffic during the TV ad.

Once the uplift has been determined, you need to compare this uplift to the cost for the TV airing to adjust for the differences in airing prices. The adjusted uplift serves as the baseline data point for optimizing your media plan. You can compare different spots, different airing times, different channels and see which works best for you.
There are a few caveats you should be aware off:
1. Not all viewers immediately engage with your website. The branding part of the TV ad, i.e. viewers recognizing your brand, cannot be measured with this method (advertisers usually use day-after recall surveys for that). Making a simplifying general assumption that non-engaged viewers behave the same across different spots, airing times and channels helps to take this out of the equation.

2. There can be interdependencies between airing time, spot and channel. A single comparison might not get these interdependencies – however, you need a lot of airings to be able to draw statically valid conclusions regarding interdependencies. Again, making a simplifying assumption and disregarding the interdependencies usually is “good enough”.

3. Most of the time viewers have multiple touch points with your website. Although they might immediately engage with your website (see 1.) very often this does not lead to the desired conversion (e.g. a newsletter signup, a purchase etc.). They might access your website three days later on the work computer which (unless you have a really good cross-device tracking solution) will count as a new user for you. They might click on your online ad a few days later because they were reminded of the TV ad they saw and only then want to take a closer look. Again, you will count this as a new user for you. This branding effect through TV advertising can be captured through proper attribution modelling. However, it is again “good enough” to focus on the direct effect of TV advertising for now and then go down the more sophisticated attribution modelling route later.

For further information, please read our TV Attribution White Paper.

We at wywy can help you set up your TV campaign tracking. We deliver the accurate TV airing data to your favorite analytics tool with a simple one-step integration. We support all of the most popular analytics tools, as e.g. Google Analytics and Adobe Analytics. If you do not have any analytics tool yet, we can guide you on how to set up an analytics tool as well. Our data can also be used with your business intelligence tools for proper attribution modelling.

Your contact for the US & Canada: (other countries)
Victor Castello