Expert advice on how to tame the data tsunami and catch the new wave of decision-making metrics.
We used to think it was simple: advertise your brand, wait for the phone to ring.
If it didn’t, you just needed to tell your story better, to a bigger audience. “Spray and pay” – spend more money and more people will buy it.
Sure, there were focus groups, consumer panels and (if you were really high-tech) test subjects wired
up to EEG caps, but when the internet came along the data collection began
in earnest.
It became possible to measure web traffic, click-through and bounce rates. Cookies told us what people were searching for and made it even easier to gather information.
Data, data, everywhere. An ocean of it, a torrent, a tsunami.
“Looking at data at its most raw form can be overwhelming,” says Clara Ng, Client Solutions Lead – Data, at Reach Marketing Agency. She acknowledges it’s easy to “fall into the rabbit hole of just looking at data and getting totally lost in it”.
Tidy up first
So how do you gain insight from the information you collect? Step one is taming your dataset, says Kevin Doyle, Regional Vice President – Data Cloud & AI (ANZ & ASEAN) at Salesforce.
“The nature of data is that it is imperfect, especially when you first begin working with it,” he says.
“You run across missing and incorrect values, spelling variations and other inconsistencies, outliers
that are correct or not, duplicate and redundant information, and other issues that can obscure the operational reality of what your data represents.
“Thorough data preparation can help ensure it’s accurate, complete, relevant and sufficiently represents your real-world operations in terms of volume and diversity, allowing marketers to remedy these faults in their data to get the best version of
the truth.
“The result is more relevant insights and higher-quality models. Ultimately, context is king when it comes to data, and a lack of context is a massive barrier to extracting meaningful insights.”
Ng likes to “start from the back of the process – from a macro level (asking what do I want to achieve/gain from it) – and work backwards to ensure I have the right datasets. That way, I will be able to identify the insight that I need to collect, or purchase, to support strategic plans that make me stand
out from the crowd.”
The right dataset, now what?
Well-organised data is a good start, but converting raw numbers into insight that will give you a competitive edge is the real trick.
“Look for correlations between actions and outcomes to understand what has driven those raw numbers,” says Sorrel Osborne, GumGum’s Head of Media, APAC.
“The strongest marketers will run nearly continuous study programmes to trace significant changes in those numbers to their specific marketing efforts. They effectively overlay the raw numbers with insights on improved brand awareness, consideration
and/or purchase intent, while considering any seasonality or extraordinary circumstances that
may have skewed results.”
Kevin Doyle of Salesforce says to do this, tech is your friend.
“With more data than ever before, much of it still in silos across a business, marketers must look to technology to help get it in shape.”
TVNZ’s Robert Hutchinson, GM Digital & Data, and Skipper Lomiwes, Programmatic & Data Activation Manager, agree.
“Marketers need to invest in platforms that unify their data and enable decision makers to find and organise information quickly,” they advise. “Smaller organisations may be able to get away with spreadsheets, but technology is a great accelerator, not only for insights but also for activation and measurement.”
Meanwhile, Kantar says it has just the tool marketers need.
Its Blueprint for Brand Growth, launched in April, is designed to show businesses where they need to focus
in order to flourish. It draws on a colossal amount of data collected
over 25 years.
“This is the world’s largest insights company in every market that matters,” says David Thomas, Chief Commercial Officer at NZ’s Insights Division, when we sit down for a chat at Kantar’s Auckland office. “Other people either tend to have behavioural (shopper actual purchase) data or attitudinal (research driven) data,” says Thomas.
“We have both at a massive scale, so billions of data points across those two, 6.5 billion in fact.”
But those datasets “aren’t designed to talk to each other” and working out how to get them communicating took Kantar more than a year.
The resulting blueprint is a breakthrough.
“It means we can map what people say they’re thinking about doing to what they actually do,” says Thomas.
“We can use that data to validate what we are saying about how marketing works. We can say, ‘Oh, you should be doing x, and here’s the proof and you’ll grow four times as fast if
you do x.’”
In a nutshell, he says: “The blueprint is the overarching framework that allows you to turn data into insights, into actions – with confidence.”
The cookies are crumbling
The slow demise of cookies has transferred the focus from third-party to first-party data. It’s required a mind shift and sometimes a costly switch of platform. Lack of resource can be a barrier.
“The biggest resistance to investing in data is the perception it’s a ‘nice to have’, very complicated/complex, or that it will cost an arm and a leg,” says Ng.
“Most marketers would love to be more data driven, but data platforms can be expensive, can require complex development and are constantly changing,” say Hutchinson and Lomiwes. “For example, data management platforms were widely adopted by many publishers in the past to unify cookies/device IDs and other such identifiers.
“But within five years, these platforms are almost redundant because of the shift to first-party data and customer data platforms.”
Did the industry rely too much on cookies?
Hutchinson and Lomiwes sum up the general gist of responses to this question: “Yes. But you can understand why: they’re cheap, easy to scale, and provide some measure (albeit a flawed one) of advertising performance.
“It was an easy way for marketers to justify spend, which ultimately influenced where budgets were spent.
“Conversations about data have become more sophisticated with the impending deprecation of cookies,” they say. “This is healthy for the industry: to truly understand what, how and why they’re buying certain inventory instead of relying on cookie-powered metrics that are unreliable and flimsy at best.”
In some ways, says Salesforce’s Doyle, life without cookies is a reminder that the job is the same as it ever was.
“Since I took Marketing 101 at university over two decades ago, the goal of the CMO hasn’t changed: to deliver the right message to the right person at the right time to help influence a decision.”
…Only now it can be done more effectively.
“By moving to a first-party data approach, primarily using data collected directly from your own audience, marketers can better understand and engage with their customers effectively,” says Doyle.
“It allows you to retarget and nurture your audience based on their preferences and behaviours, leading to improved marketing and decision-making processes.
“Marketers now can get a much richer, more accurate picture of customer behaviour, that also respects customer privacy while delivering better experiences. It’s a win-win.”
This realisation could be why, as Ng notes, “marketers have been quick to adapt, adopting first-party data to unravel invaluable audience insights” to support targeting and drive growth.
There’s clear agreement that the old methods of measuring are no longer relevant.
“The old attribution modelling, which requires cookies, that’s last decade stuff really,” says Thomas.
Phil Townend, CEO of data solution provider Magnetic, puts it more bluntly: “The demise of the cookie means that attribution modelling as it stands today is dead in the water.
“If you think about the way media is planned, we’re still using reach. We’re using digital metrics like view-through rate and completion rate to work out how effective a different channel or ad format is.
“Should I put more into channel A or B? Oh, look, I’ve got more of a dwell time on this ad on channel A than channel B. Let’s put more into channel A.”
But digital metrics can give a false impression.
“I’m looking at the screen and an ad pops up and the counter starts ticking: 1, 2, 3, 4.. But I can be doing this [looks away] or I can be doing this [looks at phone]. Or I can be looking at the newsfeed but not at the ad. A lot of the signals currently being used to quantify performance of ads aren’t correlated with real attention.”
Townend says the industry needs to move beyond these proxy measures. The good news is, he says, “the data now exists to move beyond looking at gross reach to net attentive reach, and it will take into consideration all the different attention of different channels”.
The next wave of data signals is here, he says, but the question is how we systemise them. He says contextual targeting is the way forward – and the data exists for marketers to do it.
“How do we use these new signals – like attention, emotion, psychographics, contexts – and put them into an engine that allows us to start making decisions based on those proper data points?”
Context is crucial, Townend says. Contextual targeting moves beyond the “who’s looking” of cookies, to “what are they looking at”, and stitching together rows of data attached to the email addresses of real people.
“This is why advertisers are investing heavily in customer data platforms and why they’re also investing in programmes where they can – at a really basic level – run competitions or incentives for people to give them their email address,” Townend notes.
“Once they’ve got your email address in the CDP, you have a first-party data row. But how do you then work out what their age is, or who they are? How do you add extra columns of information against that line item?”
Catch the next wave
Technology that can stitch an email address to other information and provide a fuller picture of that person will be key.
The next step is understanding their interests, so you can target them contextually rather than just demographically.
“So it’s moving from reach to attention, and moving from demographics to psychographics and context, and moving from being overly salesy to more emotive,” says Townend.
Let’s come back to Kantar’s Blueprint for Brand Growth for a minute. It instructs companies to predispose more people to their brand by being “meaningfully different”, Thomas explains.
“There is one train of thought in the marketing world that all you’ve got to do is tell people about your brand: drive salience, awareness, mental availability.
“According to our data, that explains about 40% of brand predisposition. Which leaves 60%.”
‘Meaningful difference’ explains that 60%, he says. It’s why shoppers will choose brand A over brand B, even when both are sitting side by side in the supermarket – the same pack size at the same price point.
“Meaningfulness is providing something meaningful and relevant to a consumer – at both a functional and an emotive or an affinity lens: you’re getting what you need and you’re feeling good about it,” says Thomas.
“As well as the different part, where you’re saying that your brand is unique, it’s setting trends, it’s got momentum, so what you’re getting from that product or service is different to what other brands are giving you.”
The human touch
Townend puts it this way: “If you want to build a long-term relationship with someone and nurture them and have them think about your brand over other brands at point of purchase, you’ve got to be able to understand them at a human level.”
Ironically, the robots might help us understand the humans better and faster than ever before.
Townend works with predictive analytics. “The machines can now analyse a piece of creative and predict where eyes are going to go on that asset with 95% confidence bounds,” he says.
While we talk, he uploads an advert for cheese to an AI tool. It takes about a minute, but no more than two, to analyse the creative and show a heatmap of where it predicts eyeballs would linger.
That’s the kind of information it would take a couple of weeks and cost tens of thousands of dollars to gather using real human subjects. But the AI is using 10 billion data points previously gathered from actual people – and it’s so accurate these days the AI predictions and real-life test results are indistinguishable, says Townend.
It means creatives can finesse the content to elicit more of the desired reaction, and make better use of consumers’ attention.
“With some of the next wave of predictive analytics, you can upload a video into the cloud, put your target audience – let’s say mums in New Zealand – and it will show you on a scene-by-scene basis exactly the emotions that video would make someone feel.
“So a scene of children with a cricket ball, that would make 38% of the panel feel delight at an eight out of 10, and then in second 12 of the video ad it would make someone feel like this.
“Then, if you’re a client, you can go, ‘Oh my gosh. Without even having to test them on humans, I know that among my target audience of mums, my videos are going to be driving inspiration, pride, and delight – and these are the scenes that are driving those emotions.’
“And then I can edit my videos for different platforms, so I can make my shorter edit for Facebook and I can make my longer edit.”
It means that, as well as pinpointing your audience, data can show you where they’re looking and how they feel, moment by moment, Townend says.
AI… the marketer’s friend?
If the AI analysis says a more risqué ad will resonate with an audience better than something more “middle of the road”, it might even lead to agencies and clients being more adventurous with their creative, says Townend.
But the machines aren’t about to make all the decisions yet – or lead to every advert following the same formula and becoming boring. The personal touch will be crucial.
“How do we avoid everything looking the same and feeling the same? It’s because we interpret,” says Townend. “And we, in our usual human way, add the sparkle to what’s coming out.”
Perhaps it’s all about story after all. Same as it ever was.
Catching eyes might be harder than ever, but the tools exist to help pinpoint your audience. Data can help us find out when they’re looking, what they’re into and – even better – when they’re paying attention.
This was first published in the 2024 June-July NZ Marketing Magazine issue. Subscribe here.
