25 SEP 2019
What does personalisation look like in a privacy-first world? MediaCom's Liam Brennan reflects on Google's vision of the future
The global programmatic industry is now worth $18.4 billion, and although it has matured dramatically over the last five years, it is still grappling with issues around inventory quality, transparency and the use of personal data.
But now there’s a serious challenge. The infrastructure that powers the ecosystem – the third-party cookie – is dying. The growth of walled gardens, blocking of cookies at the browser level (ITP) and a shift to mobile browsing in-app are creating cracks in the system. All this combined with the impact of GDPR/CCPA, a negative ICO POV on the industry and an increasingly privacy savvy consumer.
At DMEXCO 2019, vendors, tech suppliers, brands and industry bodies came together in various forums to highlight these issues and come up with the solutions to solve them. One presentation, from Google, predicted what the future of personalisation – and a world increasingly without third-party adtech – looks like.
“Effective Advertising and User Privacy go Hand in Hand”
Google’s “Making the web work for everyone in a privacy-first world” keynote, given by Matt Brittin (President of EMEA) and Emily Henderson (Head of Media EMEA), outlined Google’s vision for a world that’s moving from the ‘individual’ web to – or rather, back to – a ‘contextual’ web.
Google claims that this is a new world which benefits the user, the publisher and the advertiser, and allows for personalisation of advertising (which is more contextual than individualised) while respecting user privacy.
Within the keynote, a case study from The Guardian showed how its first-party publisher data can be combined with machine learning to scale. The case showed how an ad for the Google Home Mini was ‘personalised’, pulling in ingredients from the recipe shown to be used as a demonstration of voice commands for the device.
Although results from this specific test weren’t shared, Google reported a 220% uplift in ROAS (return on ad spend) from activity run from new ‘privacy first’ ways of working – an impressive number that underlines the importance of context in advertising and the benefits of data-enhanced execution.
There were uplifts for publishers, too. Increased performance means buyers are more likely to place higher value against the placements and pay more for the impression. The removal of the third-party tech chain means they – and Google – will also likely make more revenue.
But while this is a smart workaround, what these ‘personalised’ ads do not fix are the issues around digital walled gardens – the ability to move data in and out of the Google system and obtain a full view of the customer.
What’s more, while this approach might help larger publishers, Google won’t be able to directly partner with every small publisher in the world, nor will every publisher have The Guardian’s strong development team.
In my opinion, this reflects not just a power shift to Google in the post-GDPR adtech world but a white flag to privacy advocates and a step back in terms of digital capability. It misses out on a wider (and privacy-friendly) data opportunity.
A people-first web isn’t just contextual
Smart planners are already moving beyond retrospective, and often ineffective, audience-based targeting towards other data points that can better indicate what mindset or lifestyle stage a consumer is in at any given time.
Importantly, few of these opportunities use personal data, so are not impacted by the changes being made in the ad-tech ecosystem. Examples of these data points include:
Behaviours and Past Actions – Behavioural-based targeting comes from the era of contextual targeting and is still highly effective. Recently, we have been using media signals to better map the consumer and re-direct to the right destination/purchase. For example, for Bose we used search behaviour to map audiences from price-sensitive to premium-benefit-oriented and tailor messaging directionality, resulting in a 56% increase in in-store revenue driven by search.
Emotions and Moods – The emotional state of a consumer can greatly affect their purchase decisions, particularly those made on impulse. We have been working with a large CPG brand to target consumers in heightened mood states known to drive purchase (using messenger apps), which led to a 3.2% increase in the likelihood of purchase, with a 55% reduction in media cost.
Moments and Leading indicators – A person looking at buying a house, or indeed a recent home purchase, can be a leading indicator for a variety of product sells, from cars to cable packages to pet care. This is not simply a case of finding new data points, rather it’s an evolution in how we both identify and communicate with growth audiences
While the death of the third-party cookie will impact advertisers’ ability to execute, the damage is minimised by evolving from a one to one targeting at scale mindset. These ‘humanised’ opportunities, like Google’s, can come from first-party publisher data, but more often come from smarter planning and execution – and they produce solid sales uplift.
Going beyond a Band-Aid Solution
Google’s vision is a tangible solution for a cookie-less world but there are several other options that brands can take now to continue to make the ad-funded web work in a privacy-first world.
For example, they could build more first-party relationships (sponsorships, data partnerships and exchanges) with publishers, app developers and SSPs – or indeed drive first-party data collection if that provides a tangible business benefit within cost.
Or they could support ID consortiums – independent IDs seeded during the bid process that sit across exchanges for a richer view of consumers and more accurate targeting.
They could even work with programmatic ad networks to reach audiences at scale across multiple sites (often at a greater scale than the duopoly). These are valid, scalable solutions that respect privacy (although their impact is often limited to the 20-30% of inventory that sits beyond the walled gardens).
The DMEXCO conversation is often tech talking to tech – tech-based problems are met with tech-based solutions. Solutions are often replications of what has been done in the past, rather than re-inventions which can truly move the industry forward.
By going beyond programmatic band-aids and targeting consumers based on behaviours, emotions and moments, as well as looking at the context, brands can continue to market to people at scale well beyond the Google ecosystem.