Why Targeting Needs To Evolve

Long story short: ad targeting has reached a bit of an impasse, but will live on in new ways.

A brief history of targeting

Targeted advertising has evolved over time, driven by technology, data, and more recently, policy. But, like many things in advertising, targeting is about to return to its roots – well partly, at least. Before digital media existed, ad targeting was made possible through audience research (socio-demographic data, interests and purchase habits), ideas about context (like placing a travel ad in the travel section of the newspaper), time of year (running financial services ads for ISAs in March before the tax year is up) and time of day (scheduling a radio ad for chocolate bars just before elevenses). I remember being irrationally excited back in the last century when, at Mindshare, our client Shell did the first geo-targeted online advertising campaign in the UK, using IP targeting. That was the first piece of online advertising that wasn’t global.

Tech-driven targeting raged forth

Things moved on, and the ability to target advertising with ever more granularity increased, spurred on by globally scaled datasets gathered by the major tech platforms, some of which pervade the ecosystem, and first party data gathered by brands. Adtech stacks and applications were built to enable ever more sophisticated targeting, usually in the service of performance advertising. All this meant that campaigns could be optimised in real time, based on whatever behavioural measure the brand or agency thought of as important.

The debate about brand vs performance rages on

There continue to be endless debates about what proportion of campaign spend should be brand vs performance. Overall, the accepted wisdom settles at around 60% brand/40% performance, but it depends on the category and the message - and marketers need to ensure the right creative quality and brand cues to get the most out of integrated campaigns. Relying on assumptions made around smaller datasets may not be a sensible strategy. Some brands over-invest in performance marketing and focus too much on KPIs like ROAS, and this can lead to a strategy of targeting people who would have bought the brand anyway. So this is driving efficiency rather than effectiveness, which can undermine incremental growth potential.

Ultimately the answer depends on the category, the brand, its objectives, its maturity level, and evidence and predictions of effectiveness will be provided by unified modelling, based on high quality datasets and continuously measuring and optimising effectiveness.

Retargeting needs a radical rethink

Meanwhile retargeting, most often used by retailers, has a bad rep, and deservedly so in my opinion. Often cookie-driven at the point of purchase, it drives the ‘exit pursued by a mattress’ rage that comes from seeing the ad multiple times even after the item has been bought. It’s one of the parts of the ecosystem that should be happily cancelled. It’s like one of those people that talks too much – not as clever as it thinks it is. Retargeting could be cleverly used in brand advertising in the service of sequential communication, but there are very few examples of that, and they mostly exist in the movie Minority Report.

Targeting headwinds are upon us

And, to adopt the business speak du jour, we have headwinds: the first are the overall macroeconomic issues of recession, inflation and declining growth in adspend on some digital platforms. Added to that we have data privacy, regulatory developments, fines, legal challenges, and the long-awaited third party cookie deprecation, making it more difficult to collect and use data at a personal and even aggregate level for targeting advertising. And some users adopt adblockers to protect their privacy, while others don’t care. Honestly, you would think it was a bunch of nails in the coffin, and the industry would take a deep breath and return to the sun filled meadows of broad reach campaigns. But ultimately, most brands will need to adopt some form of targeting in every campaign - from the broad to the narrow.

What will happen with targeting in 2023?

Targeting will live on, of course. Here are 5 developments we will see this year.

1. Retail as a big fount of targeting wisdom: accompanied by the growth of retail ad platforms, arguably creating new walled gardens with ever more indepth data. Other media channels like CTV and DOOH are fast becoming part of the programmatic universe.

2. Realisation of the importance of creative testing: driven by data - so marketers can understand how different contexts drive short and long term impact. Individual brands need to identify the specific platform contexts which resonate best with their own brand personality. They also need to find the ad formats which best work for their specific ad messaging.

3. Behavioural data used more strategically: there will still be some reliance on behavioural targeting, using all the available digital data – but marketers need be careful to understand the correlations between behaviour and outcomes.

4. A reliance on client data: inevitably there will be more emphasis on using clients’ first party data combined with trusted third party data. We will also see more direct integrations between the platforms and targeting capabilities at the start of the process, and for effectiveness measurement, during and post-campaign.

5. Reach can be achieved in new ways: now that the bigger platforms can achieve substantial reach, marketers can use broad targeting to connect with new audiences across multiple digital channels like CTV, DOOH as well as online, rather than the narrow, more costly targeting to eek out lookalikes. So it’s partly back to the broader socio-demographic days.