As behavioural targeting dies, contextual returns to the fore: But what the hell is it?
Contextual marketing is the new black - or the new old black. But nobody can land on the same definition. As marketers scramble for privacy-compliant ways to manage scaled campaigns, what does 'contextual marketing' actually mean? Mi3 asked six experts about how context plays out in a cookie-less future – and what it means for marketers.
What you need to know:
- When it comes to contextual marketing and contextual targeting, there are as many ideas as there are marketers.
- Menulog's Simon Cheng says there is always a danger that marketers will try and overcomplicate things. It's right place, right time.
- MiQ's Jason Scott thinks contextual can still be targeted – just not tracked.
- The Trade Desks' James Bayes agrees that the end of cookies does not mean the end of targeting. Binary arguments, he said, should never apply.
Contextual targeting is back in vogue. The difference now is that we’re in an era of technological advancement.
When Menulog needed to place its Snoop Dogg ads, it chose, among others, to target music lovers on Youtube.
"Nothing there is rocket science," Menulog's CMO Simon Cheng said. "All our buys are contextually-targeted. We bought in many different environments."
For years, marketers have been chasing audiences around digital platforms on a cookie-fuelled rampage. But now, contextual marketing is making a comeback. It almost feels civilised.
The deprecation of cookies, tightening data privacy, and incoming mobile tracking changes brought in by Apple’s iOS14.5 update mean one-to-one targeting is becoming harder to do for marketers and, for the consumer, less palatable.
But in a broad canvas of industry figures, there are some soft edges to the definitions of contextual targeting and contextual marketing. Nobody has exactly the same definition.
"Marketers have a tendency to overcomplicate the simple," per Menulog's Cheng. "In the old days, it would be placing a recruitment ad in the recruitment section of the newspaper, rather than placing it in the general news."
Telstra Digital’s Executive Director Jenni Barnett told Mi3 last month that contextual targeting is yielding better results and opportunities to sell without overselling. Customers are more open to service messages over sales messages, she suggested – with conversion rates massively increasing when its marketers get the balance right.
“Get those service communications right and customers are more receptive to [an occasional] contextual sales message, more open to that conversation,” said Barnett.
“It’s value versus volume – and when we approach our customers like that in a very contextual way, our conversion rates are five times higher.”
So what exactly is contextual targeting, and how does it differ from behavioural and surveillance targeting?
The technology was so interesting, challenging, and complex, and there has been this excitement with big data and speed and real time – we got carried away with the toy, to be honest.
Contextual based on content
Integral Ad Science Country Manager Jessica Miles says the definition is simple: contextual marketing is about where the ads appear.
“Contextual targeting is back in vogue; it’s having a bit of renaissance. I used to work in mobile and those individual signals from cookies weren’t as prevalent on mobile, so we had to use contextual,” she says.
“The difference now is that we’re in an era of technological advancement. We have more enriching signals than before. Some signals are words on the page, images on the page, the context of the content on the page. Some of the things that are important are this idea of sentiment and emotion, being able to discern the way an article can be read by a human being.
“Take the Royal Commission into the banking sector, for example. You should be able to discern the sentiment, and that could make a difference to whether a brand wants to be associated with that content.”
Contextual targeting is black and white, Miles says. It doesn’t collect a user’s past data in any way, shape or form, making it future-proof in terms of data and privacy.
“It’s determining the signals on a page,” she says. “If you’re targeting people looking at buying a house, it might not just be appearing on a real estate page, it might be mortgage comparison sites – everything that could relate to buying a house from a contextual point of view.”
Think Premium Digital General Manager Venessa Hunt argues context has been the basis for all marketing since marketing began. It wasn’t until the rise of digital that the first opportunity to target specific people entered the market.
“[Contextual targeting] uses the content people are viewing/reading, someone’s interest or factors like location to target advertising to them and logically provide a better outcome for a clients marketing dollar,” she said.
“With the depreciation of cookies and the recent tracking opt-in changes from operating systems, contextual advertising in digital, in effect, begins to mirror the rest of non-digital advertising, utilising the same principle of finding an audience based on preferences for particular content or location. We use this logic in all other advertising channels.”
It’s not a perfect example, but if I’m lying in bed and it’s midnight and markets are closed and I’m getting served a ‘trade stocks now’ ad, the time and place is off. That’s how I think about different data sets. I would argue all of this makes up contextual targeting.
Contextual based on life stage
If ‘contextual targeting’ can be understood as ads displayed where the user currently is – rather than where they have been, the term ‘contextual marketing’ is another story.
Roy Morgan Research recently partnered with EngageTV, the connected TV (CTV) platform from Switch Digital, and will be able to target advertising to specific postcodes. Roy Morgan’s ability to segment the audience, based on thousands of interviews, is included in the broader picture of contextual marketing.
“There has been a move away from a truly consumer-centric perspective,” according to Roy Morgan Research CEO, Michele Levine.
“The technology was so interesting, challenging, and complex, and there has been this excitement with big data and speed and real time – we got carried away with the toy, to be honest.”
Roy Morgan Research centres its segments around 54 Helix Personas split into six “communities”. These personas aim to define all parts of the Australian population and give context and structure to ad spend.
“I think the market is becoming more sophisticated and more understanding,” suggested Levine.
“There’s another movement that’s cutting across this – an awareness of privacy, at a legislation level, at a corporate level, as they become more concerned, there’s also the feeling people are becoming more aware of how their data is being used.”
The winners in the post-cookie world, Levine believes, will be premium media players.
“Those with addressable audiences, those that have a mechanism to communicate – they will do well.”
Sammy Bolton is the Managing Partner at media insights consultancy Audience Precision. She thinks getting away from the 'creepy' aspects of digital advertising is a good thing for everyone.
“It's probably not that different to what we did a long time ago. It's going back to where the consumer is and what they are into.
“It’s about the environment. It’s not targeting a ‘metro man’ wherever he is, which is what third party cookies are doing now. It’s about choosing contextual relevancy based on segments of the audience and the product.
“If you’re trying to sell a car, you might go across all car stuff. Contextual is focusing on content that you’re going to advertise in or around. Third party marketing is building an audience and, wherever they show up, you’re going to target them. You follow them around the internet. It’s gotten so stalker-ish on the internet.”
We’re inherently wired as an industry to want to make binary statements that one approach will win over the others.
Contextual based on data
Whereas once upon a time, contextual targeting was a car advertisement in an auto magazine, now the industry has access to vast amounts of de-identified data, according to Jason Scott ANZ CEO at MiQ.
“There’s a lot of nuance out there,” he said.
“In a digital sense, contextual targeting has evolved massively since its early days. The simplest way to think about that is it’s no longer just about targeting pages that talk about holidays. It can go way beyond that. Think of it as a framework that surrounds an event.
“The way we would think about that, two things: we don’t look at one contextual dataset in silo.”
Rather, time and geolocated place are elements that can enhance a targeted contextual offering.
“It’s not a perfect example, but if I’m lying in bed and it’s midnight and markets are closed and I’m getting served a ‘trade stocks now’ ad, the time and place is off. That’s how I think about different datasets. I would argue all of this makes up contextual targeting," said Scott.
“There’s global industry recognition that the way consumers were tracked in the past is no longer fit for purpose under new privacy concerns.
“If you’re a marketer, agency or anyone involved, you’re going to need to navigate multiple datasets: authenticated data, first party data, anonymous data – this is included in contextual marketing.”
Think Premium Digital's Venessa Hunt believes contextual advertising can deliver results alongside more reliably “safe” content.
“The next frontier of digital advertising will be won by those who have highly engaged (and safe) content,” she said, “with a strong first party privacy first data strategy that underpins the context in which an ad is seen.”
Not dead yet
However, some believe that reports of the death of behaviour targeting are greatly exaggerated – and that it will not die off with third party cookies.
The Trade Desk’s James Bayes said marketers will have to be across a range of approaches: contextual targeting, “cohort targeting” – like Google’s FLoC – and cookie alternatives like Unified ID 2.0 – an initiative being driven by The Trade Desk and others
“We’re inherently wired as an industry to want to make binary statements that one approach will win over the others,” said Bayes.
“But the reality is more nuanced and it’s likely a range of approaches will prevail. What’s obvious right now is that not enough people have a deep understanding of the consequences of the change and the alternatives available to them, and have a clear and specific plan to address it.”