Consumer analytics firm ViewersLogic has passively measured consumer behaviour 24/7, including all their TV viewing (content and ads), online behaviour and offline store visits.
The analysis on the impact of TV advertising on online platforms was conducted in sectors such as supermarkets, automotive, travel, insurance, financial services, lottery, gambling and online clothing, which represent a cross section of products associated with both impulse and considered purchases.
Marketing Gazette recently caught up with the company’s CEO, Ronny Golan, to discuss what marketers can learn from the study’s findings…
What motivated you to analyse the impact of TV advertising on online platforms?
The goal of the ViewersLogic’s project was to enable companies to understand TV’s effect on advertising. Today, when people advertise online, they know exactly how their money performs, including the cost of acquisition and purchase.
However, when companies go to TV, it’s much more difficult to keep track of their performance. TV advertising has experienced a significant decrease in recent years; thus, it accounts for less than 25% of all ad expenditure in the UK. Nowadays, companies usually have cross-media campaigns, including TV, social media and billboards. In this case, if the sales go up, this is good, but the question is ‘which one causes it to go up?’.
Until recently, no one could measure this, and our goal was to find a solution to this problem. Here at ViewersLogic’s, we wanted to really understand the effectiveness of TV advertising.
Although the users’ ‘journey to purchase’ is complex, nowadays media is measured individually. One analytics company usually looks at online advertising, another at TV and third at social. Therefore, current technologies see only parts of the ‘journey to purchase’. However, the only way to really understand consumers’ behaviour is to completely change the way of measuring it, thus, instead of measuring media, to start measuring people. This means that the same person should be measured across TV, online and the real world.
Single source data is the only way to understand the whole journey.
How did you gather the data and conduct the research?
We had a panel in the UK and each of our panellists installed an app on their phone, so we measured the participants entirely online. The app works passively on the background of the phone 24/7 without the user needing to do anything. The app speaks with the TV to get information about what people are viewing on the TV. We also get information on everything people are doing on their phone, tablet and PC. We know which store they are entering in the real world. Having these three things is the most comprehensive consumer data set.
Here at ViewersLogic, we spent five years on building our technology and collecting the data. The analysis was conducted in sectors including products connected to both impulsive and considered purchases, such as supermarkets, automotive, travel, insurance, financial services, lottery, gambling and online clothing.
What do you think were the most interesting results of the project?
One of the most interesting and surprising findings was the effect of frequency of TV advertising. Consumers who clicked on a Facebook or Google ad of a certain brand, watched on average 39% more TV ads from the same brand in the week before than users who did not click on an online ad or visited advertiser’s website. This effect was the highest on Facebook clicks and we found that consumers who clicked on a Facebook ad saw 48% more brand specific TV ads of the same brand compared to users who didn’t click on a Facebook ad.
Users who interacted with the website of a brand were exposed to an average of 28% more ads and consumers who clicked on a Google ad watched an average of 42% more TV ads.
After seeing an ad for the first time, people won’t do anything, but when they see it after five or six times, they’re more likely to take action. This means that companies need to get to a certain frequency threshold in order to get people do something. Thus, the successful campaign is not only about getting the highest reach, it’s also about required frequency.
In order to understand the whole journey of purchase, companies should go beyond the last click model. If I see a Facebook ad and click on it, Facebook will claim 100% contribution. Therefore, the company will see that 60% of the traffic comes from Facebook and 40% from Google. Then, the question is ‘why should I put money on TV, when TV drives nothing’? Our data reveals that people who clicked on online ads didn’t do it only because of the Facebook ad, but because they’ve seen the TV ad. In this case, companies need to give 50% to Facebook and 50% to TV, or maybe 20% to another online source. Measuring more than one media is the way to understand the whole journey of purchase.
Which were the best performing sectors and why?
Gambling and online clothes were the most effective sectors to use TV. Users who clicked on online ads saw 68% and 47% more TV adverts, respectively, than those who did not take actions. The reason for this might be that they have a lot of experience in using TV advertising. Another explanation is that they are always on and they advertise so much that they get to this frequency, where they pull people to take actions.
For example, we looked at people who visited the Coral website and we measured how many ads they saw in the previous week. Then, we looked at people who didn’t visit the website, but because gambling is a niche industry, we didn’t look at all people who didn’t visit the website. We looked at the gamblers who didn’t visit Coral. In this way we excluded people who are not likely to click on the website no matter how good the campaign was. What we found was that, if we look at the same people who visit the website and those who did not, it’s possible to isolate the effect of TV on the site visits, Facebook clicks and Google clicks. As a result, gamblers who clicked on Coral’s adverts or visited their website were exposed to 117% more TV adverts than other gamblers. When we measured this for the online clothing sector, for example, Jamaco, the result was 85%.
This data shows that there is no better explanation than that TV advertising fuels Facebook and Google advertising and secures its position as the number one medium at brand building and driving online traffic.
What were some of the biggest challenges you faced with this study?
One of the setbacks we experienced was that all this data needed to be collected without affecting the user’s phone, like slowing it down. We had to develop the ability to connect to the TV and get information from the TV without opening the microphone. We had to make sure that there is no data that could be problematic for the user. Then we had to develop something that enables us to collect the online data from the phone, PC and tablet, as well as, to know which store the user is entering in the real world.
We were also using club card data to know the purchases in the real world. As a result, we were getting more than 2 billion data points per day. That’s why it took us a long time to analyse and gather comprehensive data to get a clear picture of consumers’ behaviour.
How do you apply the data to your work at ‘ViewersLogic’?
We created a dashboard for our clients to enable them to know exactly how much each channel costs to bring a person to their website, to register or buy something from the website. We have the data on every company in the world and the dashboard shows not only how our client is performing, but also their competition. They can see what works for them and what doesn’t, the effect of frequency and effectiveness of competition ads. This really helps companies to understand how much money they need to move from channel A to channel B in order to optimise their performance and TV campaigns.
Our method is very different from the five-minutes attribution model used today in the market. Other consumer analytics companies are calculating the baseline of the website (how many people are supposed to be on the website, if there are no TV ads), and then they attribute all the extra visitors to the website to the previous ad that was aired in the past five minutes. However, only 1% of the traffic comes within five minutes and companies cannot optimise their campaigns according to this 1 %. The main shortcoming to this model is that it does not take frequency into account. If a consumer sees an ad on a certain channel, it’s not the same whether it’s the first or the fifth ad. That’s why our attribution window looks at all ads the user has viewed in the week before entering the website. This means that if the user viewed four ads on four different channels, we’ll give each channel only 25% of the attribution.
How do you think marketing trends will change in the future, bearing in mind your data?
I believe that one of the main problems nowadays is that TV and online advertising are treated as completely two separate things.
When companies are doing TV and cross-media campaigns, there are currently two strategies to optimise their campaign. The first one is when companies aim to optimise their reach, which means that they want to reach online people who haven’t been reached on TV. The second strategy is to reach a certain number of people on TV and reach the same people online in order to activate them. This is the better strategy because people should see the TV ad five or 10 times and when they see the ad online, the chances to click on it are way higher than, if they haven’t seen it on TV. This strategy is more effective for online companies and our technology can help them design their campaigns in a better way. We believe that the campaign should integrate TV way more than just online and this will be the future of marketing.
If a company has a campaign with an 80% rating, is it successful? The answer is that the fact a company has an 80% reach is irrelevant to the success of the campaign. The same is with online advertising. Companies moved from CPT (cost per thousand – paying for 1,000 people to see the ad) towards CPA (cost per action – paying only for people who take a specific action) in digital advertising.
With our data companies can understand that TV advertising increases the impact of online advertising and they can start using CPA with TV advertising. Therefore, the main impact our data will have is combining TV and online advertising to optimise campaigns, making them more effective and less costly.