Single Source Data: Explained

Tim Cross 05 January, 2022 

Last year we finally saw some good progress on some of the industry’s cross-media measurement initiatives – NBCUniversal’s CFlight gained further traction from broadcasters, while the WFA and ISBA claimed success in crucial proof-of-concept trials for their own efforts.

However even when this measurement ‘North Star’ is reached, ad measurement will be far from an exact science. These cross-media measurement solutions will help marketers better understand exactly how many individuals saw their video and TV ads, and how many times. But it won’t give a full picture of all the different ads these individuals have seen across all media formats, and it won’t show which ads were most crucial in driving conversions.

Some measurement companies believe taking a fundamentally different approach to measurement, and focusing on ‘single source data’, can solve these problems. 

The Basics

Single-source data is essentially the measurement of marketing exposure and purchase behaviour over time, for the same individual or household.

So it’s akin to a traditional TV panel, in that consumers agree to have their behaviour tracked. But while a regular TV panel would only monitor TV viewing for participants, a single-source data panel would also measure as much other media consumption as possible – from online ads, to outdoor, radio.

And single-source panels also track buying behaviour, to show when consumers have actually gone on to buy products which they originally saw in ads. They will also measure other outcomes that advertisers may be targeting, such as registrations, app downloads, etc.

The Technical Details

Single-source data often essentially combines a host of existing measurement techniques for different media. For example, TV viewing may be measured using automatic content recognition data, phone location data might be used to track which outdoor ads an individual has seen, and cookies or other online trackers can be used for web browsing.

But what’s unique is that all of this data is tracked and tied together for individual consumers, meaning that an individual consumer’s entire purchase journey can be seen.

Some single-source data initiatives have created wholly new technologies to facilitate data collection. Some, for example, require panellists to carry a device with them which monitors audio, which is analysed to deduce which ads that individual was exposed to.

Since it is still a panel-based approach, any conclusions around reach and impact of ads have to be extrapolated from a sample of the total audience.

But while other measurement techniques may try to tie together audience-wide data from different mediums to model consumer journeys, single-source data, when it works well, shows actual consumer journeys.

There is variation however in how the data is used. One variable is the total timeline which marketers look at. An effective ad can help drive a conversion weeks, months, or even years after it’s been seen by a consumer. But ads which have a significant impact years after they aired are few and far between – most marketers will want a cut-off point somewhere. It’s up to the marketer how far back they want to measure.

Another variable is how exposures are weighted and combined via an attribution model. Seeing the whole consumer journey still doesn’t tell the marketer how big of an impact each exposure had, so the marketer must decide how to adjust weighting based on medium, passage of time, and other factors.

The Pros and Cons

Proponents of single-source data say that it makes measurement much less siloed between different media channels, and gives a fuller picture of consumer journeys.

The fact the user behaviour is measured over time means that channels less suited to direct response aren’t discounted, as they might be using other measurement techniques. For example the impact of a TV ad can still be measured, even if the consumer doesn’t see an ad and then instantly go to the advertiser’s website to buy a product.

Single-source data in a sense combines the privacy-compliance and accuracy of  panel-based measurement with the ability to measure across different media channels.

But it can be expensive and difficult to do. Several attempts at collecting single-source data, including an effort by Nielsen, have failed, with cost cited as a primary reason. Technological advancements are helping reduce costs, and making it easier to record user behaviour in ways which don’t place a big burden on panelists.

Despite this, data collection still isn’t perfect – not all media consumption can be easily tracked. And single-source data is only as strong as the attribution model used – meaning it’s still susceptible to erroneous conclusions in the wrong hands.


About the Author:

Tim Cross is Assistant Editor at VideoWeek.
Go to Top