Nordic TV operators can capitalise on the first-party data they already possess to drive TV advertising revenues in the region. The data’s potential lies in its much higher granularity than existing sources, and this enables it to tell marketers who watched which programs and ads, and for how long. First-party data will help to add accountability and targeting to a medium that is already the best way to achieve reach, and one that most people each day watch (as the below graph, for Sweden, shows)
TV viewing measurement today
Viewing data in Sweden is collected by MMS (the equivalent of Nielsen in the US). MMS provides live reports on both traditional TV viewing and online viewing using people-meter technology that listens for audio fingerprints associated with programs and ads and then determines what the audience has watched, for how long, who is watching, etc. MMS uses a panel of 3,000 households to represent the Swedish population’s viewing, and marketers rely heavily on the data from MMS. In Denmark Kantar Gallup uses 1,200 homes in its panel.
Panels generally do an excellent job of estimating how many people, and who, have watched the main shows and channels. Panels aren't always as accurate at measuring smaller channels, however, more importantly, they don't go beyond simple measurement. Panels may show that 3m people watched Eurovision, but they don't tell you what those 3m people watched afterwards. They also don't tell you much more about the audience, beyond basic demographics.
Enter first-party TV data
First-party data, as the name suggests, is data about viewing that comes from the TV service provider. The data can come from a Pay-TV provider, using set-top-box data, or smart TV data from a device manufacturer (multiple providers are using automatic content recognition (ACR) technology to understand content viewing on their devices).
A significant benefit of first-party data comes from the size and granularity of the data set. Any device connected to the Internet (a set-top box, OTT device or Smart TV) can produce viewing data. We don’t know exactly how many of these devices are connected in the Nordic countries, but Sweden alone has 1.1m IPTV households (which by default are connected), and another 3m cable and satellite homes, of which we can assume a significant proportion are connected.
The value of this data comes in that it is capturing all viewing, regardless of the source. Despite the hype about OTT, the average person in Sweden still watches 80% of all their content via traditional TV, as the below data from IIS shows.
Many service providers already collect this data for themselves and use it internally, but in the Nordics, most do not share it with third parties. But there is a real opportunity for them to do so. Advertisers would benefit significantly from a data set that is not just large, but targetable, and that can support the kind of attribution (e.g., did the person that watched this TV ad buy this product or respond to this call to action) that has been so important to the growth of digital media. TV ad inventory can be much more competitive with other digital platforms with these advantages.
Sell-side companies are not the only ones that can benefit. Broadcasters’ programming departments get much more granular detail on viewing habits, which makes their licensing and scheduling efforts substantially more effective and efficient. The result is a better TV viewing experience, something favourable for everyone in the TV ecosystem as a while.
Not all data is created equal
The quality of a first party dataset doesn't just depend on how big it is, but also on the scope, and the consent of the consumer.
We tend to split the level of data that an operator can share into three tiers:
Tier 1 - anonymised data
Data with no association to any household, and no demographics. In this case, the data will show that a specific device watched a particular show and that that device went on to watch another show, or use another app. There is also no matching of the data to third party data sets
Tier 2 - semi-anonymised data
These datasets include demographic information, but nothing that is personally identifiable. Companies typically use these datasets for audience planning and segmentation
Tier 3 - personal data
Personal data is data that includes personally identifiable information (postal code, IP address, location) and that can be used to drive user-level targeting - e.g. attribution and targeted advertising.
The information in Tiers 1 and 2 does not require consent but is less valuable to marketers. To collect data at Tier 3 requires consent from the consumer and this consent has to be opt-in to be GDPR compliant. GDPR is still in its early days, so it is hard to predict what levels of opt-in may be. But even if only 30-50% of consumers provide opt-in, the result is still an extensive dataset.
How to use the data
With many deployments in EMEA and North America, our data science and data engineering teams have helped a range of operators, broadcasters, marketers, and brands to drive value from first-party data:
Audience Indexing for media buying
We have worked on numerous campaigns where we have provided audience indexes against first-party data sets to target media buys. Much of Dativa’s current work in the US is helping the sell side set up sales tools and inventory optimisation capability.
Using TV data for attribution
First party data can be used to link a programme view to another behaviour. We do this commonly with tune-in reports: analysing and understanding if a consumer has tuned in to a program or not when they have watched a show promotion, and identifying the content that best drives tune-in. This information can be abstracted to any trackable behaviour – e.g. downloading an app, visiting a website, or buying something – via a cookie or IP address match. This kind of attribution helps marketers to understand which ad impressions were most effective in driving the target behaviour.
Cross-platform measurement and analysis
We use first-party data sets to understand the effectiveness of reaching a specific target audience across TV (linear and addressable) and digital. We work with customers to find which combination brought the most effective reach and tune-in.
Summary - the power of first party TV data
First party data has great potential to augment traditional panel-based data sources and to help grow TV as a whole. Even as an anonymised source, without any other data matched to it, it provides a much more granular and detailed view of multiplatform viewing. With demographic data, and when matched to further data sets, with consumer opt-in, it brings a robust level of targeting and attribution to TV. By combining this with the power of TV as a branding medium reach, first-party data will ensure that TV can compete with digital, and continue to remain in rude health.