The Guardian shares its programmatic strategy as third-party data disappears
Conference Blog | 23 September 2020
Most people know The Guardian for its voluntary payment model. But the trusted UK news media company is also making a name for itself in programmatic advertising.
Russell Foxley, digital director at The Guardian, shared with INMA members during a Webinar on Wednesday the different types of programmatic activation channels the company uses to reach its premium audience.
Programmatic channels: pros and cons
Foxley opened the Webinar by explaining the four activation channels through which The Guardian operates programmatic:
- G+
- Programmatic Guaranteed
- Preferred deal
- PMPs and OMPs
These channels are determined by a range of different factors including pricing, auctions, technologies, and performance. Each channel has a cost associated with it, such as ad-serving fees and content verification, but there are also hidden costs that cannot be calculated but must be accounted for.
G+ is a managed service run for the advertiser. The other channels are all self-service in which the buyers operate it for themselves. G+ and Programmatic Guaranteed are also guaranteed, while the others are not. All of the channels use first-party data except OMPs, which use third-party data.
When it comes to pricing, the more data and guarantees are offered, the higher the charge for those impressions. Therefore, OMPs are the lowest priced, going up the chain to G+ being the highest priced.
“If you start at the bottom with an open market purchase, these are often different brands looking to do retargeting campaigns,” he said. “If you are a luxury brand, you want to come in more at a guaranteed, because you are more interested in brand safety. There are technical advantages as well and blocks other competitors from advertising next to you.”
In between those two approaches is biddable advertising. “It’s always worth approaching a client, approaching an agency, asking whey they are taking their approach through each different channel,” Foxley advised. “If you can start those conversations from knowing a little bit about how they work, you’ll start off in a better stead.”
An approach that’s rarely discussed is Audience Guaranteed.
- It’s an often overlooked activation channel.
- Finds addressable audiences across publishers by utilising first-party data.
- Allows buyers to identify, target, and execute against their own first-party data segments.
- Forecasts estimated delivery against segments.
- Offers increased scale over open auctions, whilst ensuring target audience.
“The reason it’s so advantageous is because there’s no inference, there’s no guesswork,” Foxley said. “It means we’re targeting the exact audience the client wants to speak to.”
He believes the Audience Guaranteed product will be hugely important in coming years. He outlined how the supply chain works in a slide presented to INMA members in attendance:
“The reality is, all this happens in milliseconds and it’s a two-way conversation, it’s not a linear process,” Foxley said. When the customer overlays their own DMP over all this, it affects it as well.
“This as a product is somewhat short-lived, but until other products come along this is a particularly valuable product for The Guardian,” Foxley added.
First-party versus third-party data
Data is fraught with inaccuracies and fraud, or is simply worthless. Oracle, one of the biggest data providers in the world, can take a look at a user’s cookies and displays different audience segments they might be interested in. However, it sometimes backfires, providing very inaccurate information on the user’s preferences.
“This is evidence that third-party cookies don’t work in the way they were meant to be working,” Foxley said. “We’ve gone from this minor trade-off of surveillance to it being an actual nightmare. As an industry, we didn’t move to course correct.”
This has led to new technology such as browsers killing third-party cookies, legislation to ensure user privacy, and public awareness of misuse of data and privacy. This may be the year that all of this starts to change, largely driven by legislation and public demand, Foxley said.
- 41% of readers hate online advertising.
- 71% think ads are becoming more intrusive.
- 48% don’t know how or where brands use their personal data.
“Individuals are aware of the value of the data and how it’s being used,” Foxley said. All of this is leading to the death of third-party tracking.
“Now that [third party cookie tracking] is being removed from the market, there’s a huge scramble to readjust,” Foxley said. “You can’t take a new approach with the technology that’s been around for years. They are looking at the industry to find a fix that isn’t there. As we’re seeing cookies being taken out of the market, we’re seeing the impact this is having.”
Yet many agencies’ budgets are still being allocated to third-party data and are unaware of these changes.
At a browser level within the UK, Safari and Firefox combined represent 37% of the 53 million total online audience. Therefore, this is a huge piece of the market unavailable to third-party data.
“Roughly a third of the Internet has just vanished from being targetable,” Foxley said.
First-party data can plug the gap
Third-party data is broad, cheap, opaque, and available to all. First-party data, on the other hand, is pure, premium, accurate, and operates in closed environments.
The narrative that The Guardian took to the marketplace is called Guardian Origin. “It’s been particularly successful,” Foxley said. “We’re still having deep conversations, but some of the products and services are of huge interest to agencies.”
This modern data is offering best-in-class segments built on edge computing. Foxley explained that, for The Guardian, this approach is:
- Interest based: What our readers are interested in.
- Identity based: Who our readers are.
- Intent based: What our readers are up to.
“The key message is the closer you can get to the origin of the data, the more reliable that data will be,” Foxley said. The Guardian principles and Origin product combined means the company can offer advertisers something they can’t buy elsewhere:
- Meaningful reach, with huge scale and undivided attention.
- A progressive audience, with pretty much any hobby, passion, or interest the advertiser wants to target.
- Trust. With trust in advertising at a low point, Guardian Origin can help build that back up as the most trusted news brand in the U.K.
Brand suitability
Brand suitability comes down to whether it is appropriate for a particular brand to be shown next to certain content. For example, it would be appropriate for Greenpeace to appear next to content on greenhouse gas emissions, whereas a car manufacturer might not want to. It would be not be suitable to show an advertiser next to news about a terrorist incident.
“That doesn’t mean it isn’t brand safe, it just might not be brand suitable,” Foxley explained. “There’s perception that hard news is damaging to a brand.”
However, data from The Hard News Project shows readers do not react negatively to this; they are able to separate the brand advertising from the news content.
Brand suitability has several layers:
- SSP protocols: Standard ad content filtering from ad tech partners to eliminate gross negligence.
- Contextual targeting: Focus on the verticals that attract an advertiser’s audience.
- Keyword blocking: The Guardian uses a keyword list to block or target content containing those themes or values.
- Third-party verification: The Guardian partners with Integral Ad Science to ensure site content is independently screened and verified.
- Brand Council: Containing senior editorial and commercial staff, the Brand Council is active 24 hours a day to assess the suitability of breaking news for advertisements.
More brands have been relying on keyword blocking, Foxley said. While they were effective a few years ago, he believes they often are not updated and therefore become less useful.
“It’s not until we know who an individual is that they become more valuable. Earlier, we didn’t restrain ourselves and it would appear we went too far. There’s now a balancing act.”
This means there is a need for balance amongst three separate needs:
- What consumers will accept.
- What is commercially viable for publishers.
- What buyers need for effective campaigns.
Foxley believes that unless publishers can find a way to successfully balance these needs, they might be unsuccessful.
Data shows:
- Only about 10% of the Internet is authenticated. “We know who they are as an individual and they’re valuable to us,“ Foxley explained of this group.
- 50% of Internet users’ data relies on third-part cookies and that’s shrinking rapidly.
- 40% are anonymous users.
“That’s the world we live in, and I think people think it’s much better than it currently is. Were moving to a world where the majority of the Internet will be anonymised. A lot of publishers are kicking in with log-in and registration walls, but I think it’s unrealistic to expect that the majority of users will actually register and log-in to each Web site they visit. As third-party cookies decline, we’ll see a huge part of the Internet becoming unrecognisable.”
The four scenarios
Foxley disucssed the possible four scenarios of user privacy in the future.
- The Google scenario. Google owns a huge part of the Internet audience via multiple platforms and devices, and has a full view of any individual. The company is operating close to a monopoly.
- Paywalls and registration gates. However, the entire Internet can’t be restricted and put behind paywalls. This would be financially restrictive and unrealistic. Single sign-on is one possible answer, where users enable a single sign-on to log into sites, such as their Facebook or Google account. However, this quickly comes back to the monopoly scenario.
- Sign-on ID. This relies on a single company having an ID on various individuals. Also leads back to the monopoly situation where one organisation holds the keys to consumer data.
- Data clean rooms. This is a new round of technology that should solve a lot of issues with third-party cookies. With data clean rooms, a user shares their data with the clean room, it’s compressed down into pixels that are unrecognisable, yet would be available to the parties that the user approves for it to be shared with. The data can’t be reverse-engineered, allowing for privacy of the individual’s personal data.
Foxley urged news media organisations to consider the future of data privacy, and how these various models will fit in with advertisers’ and users’ needs to create a space for modern advertising.