Second screen viewing has almost become a norm of how TV is being watched these days. If you don’t trust me, google “second screen viewing statistics” and you’ll be blown away with the amount of research out there. As a marketer your ability to track TV’s impact on website traffic and conversions are essential, more so if you are an e-commerce business. Multi-touch attribution as a concept has finally gained momentum with marketers slowly moving away from the last touch world.
TV Attribution: Why Should you care about it?
Just in case you didn’t google “second screen viewing” below are some stats to convince you
- 3 out of 4 viewers are having a second screen experience while watching TV
- Roughly, 4 in 10 will search for information about the show or product advertised
With the advent of connected TVs, this argument becomes more compelling. Actual campaigns I’ve measured in the past have also shown significant increase in on-site behavior. My aim for this post, is to help you understand the current standards of TV attribution, how you can leverage this for your business and what some of its limitations are
So how does this work?
There are plenty of vendors out there and almost all of them offer a solution to measure traditional offline media. In the case of TV, the most common way of measuring its impact is to measure the site activity during a time interval before and after the spot is aired and then calculate the incremental lift resulting from it (refer below).
While this method works very well for visits, some vendors deploy the same technique to calculate conversions and revenue. This to me is a little bit of a stretch. If your interest is to do this, media mix model is the right way to go about it.
TV as a medium is meant to have a long-term impact on your brand. Evaluating TV’s performance like a display campaign is not the right approach. I have worked with attribution platforms that offer this solution and I’ve often found that the results generated have never helped with revenue optimization. So if you are thinking of signing up for that TV/offline module from your attribution vendor, think twice. Save your dollars and invest it on an analyst.
Why isn’t this good for measuring impact beyond visits?
To explain this we need to understand the technical aspects on how tracking is done on websites. Web analytics tools tracks users by placing a cookie, (No! not the one that you eat 🙂 this joke never gets old) which is a unique identifier for every user that visits the website. The users are also assigned a source of entry each time they visit the website, be it direct, search or social. The categorization in this case, happens at the individual user level. This unique ID tracks the user’s entire journey across multiple visits to the website.
On the other hand, while doing TV attribution the aggregation happens based on the chosen time intervals (5 mins pre and 5 mins post) and not at the user level. The users’ source of entry to the site would still be categorized as a direct visit or as visits from paid or organic search. Unless you use a vanity url, there is no way of accurately assigning a user to TV. If you can’t tie it back to the source you wouldn’t able to track the user’s behavior beyond that visit.
How can you act upon this data?
The ultimate goal of doing this analysis is to maximize the efficiency of your TV spending. Unlike other channels which are measured with conversion (CPA) as a goal, this analysis will help you understand which networks within your media buy are more efficient in delivering immediate impact. The analysis will also help you optimize towards the ideal day part mix, the best creative, channels that expose your brand to new users, etc. Depending on the length of your TV flight, you’d be able to make adjustments to your campaign at the mid-point or for shorter flights, you could use these learning for future TV campaigns.
I will cover this topic in more detail with actual examples in subsequent blog posts. I hope you liked this one. Please share your views and opinions in the comment section. Until next time!!