Type to search

2021: Prepping for an Attention Recession

Technology & Data

2021: Prepping for an Attention Recession


The team at Audience Group looks far beyond 2021 and considers the year 2030 – how can marketers analyse the possibilities and shortfalls of this imagined future in order to focus their efforts and prepare for what is next? An attention recession, hyper-personalisation and an algorithm for everything that you see. Welcome to 2030.

Looking ahead to 2021 is short-sighted. It’ll only get you through next year. Maybe. 

What if you could leap forward in time, take a look at what marketers will be facing in 2030, and work backward from there? You’d be ready for anything.

We don’t have a time machine… Or do we? Even if we were visitors from the future, we couldn’t tell you. It wouldn’t be safe for the space time continuum, or whatever… But we can acknowledge that the rate of technology change and digital and mobility innovation has only sped up thanks to this pandemic we’ve all found ourselves dealing with. And we can make some educated guesses as to what’s to come.

Focus for 2021: We’re going to look ahead to an imagined 2030, and consider what marketers need to start or continue doing in 2021 to start prepping for that world. 

Privacy or personalisation?

In a world ever-more-managed by AI, assumptions will continue to be made about what it is that you want to see next, driving your social and information feeds down a particular path without any opportunity to think outside of the information bubble you help to curate with clicks, likes, reads, comments and shares. By 2030 will ‘the algorithm’ control EVERYTHING that you see and hear? 

This would make it increasingly difficult for marketers and advertisers to get someone’s attention unless they are offering something the AI determines to be what that person wants to see.

Privacy concerns are likely to present significant roadblocks, as well. People want entertainment platforms to remember their preferences and personalisation on their favourite sites. They want extraneous information to stay out of their feeds. Yet they are wary of their movements and behaviours being tracked and their personal data being captured, shared and sold. Consumers find it creepy when they are fed ads that mirror their online product research, or seem to echo what they’ve just been talking about with their partners or friends. 

The possibilities are … scary? 

People talked for ages about that divisive Minority Report scene where Tom Cruise walks through a hyper personalised shopping mall with holograms referring to him by name. There are consumers who would never want personalisation – and the associated knowledge about their purchasing behaviour and preferences – to ever get to that degree. And others who think that is how it should be, or how it will inevitably be. 

It may very well be like the minority report by 2030 – at least from an advertising personalisation perspective. If that sounds farfetched, it shouldn’t. 

Personalisation is already possible to a degree that many people (who are not focused on data science and analytics) would probably find surprising. 

There are sensors and technology solutions available to deploy right now, that would enable personalised information to be presented to shoppers as they approach a display in a retail outlet. Or that would enable the activation of a hologram host to welcome a returning hotel guest by name.   

But consumers resist keeping Bluetooth and loyalty apps enabled on their smartphones or watches. They avoid scanning QR codes. They sign in to the sites they use but react negatively to targeted ads and personalised prompts. They are worried about privacy, data capture and security implications.  

It’s not lack of technology or capabilities stopping the progress of personalisation; it is a lack of user engagement and user buy-in. What will have to change in terms of society’s use of and trust in technology, data security, data privacy for that to happen?

Focus for 2021: Studies have shown that consumers are more willing to share data when they feel they are getting fair value in return. Marketers and advertisers need to consider what they need to bring to the table, in order to earn engagement and attention. What value can you add? How can you communicate that value? If privacy concerns are a critical focus for your target audiences, what can you do to reassure them from a security and data protection perspective? 

Hacking the attention algorithm   

Individuals try daily to sway ‘the algorithm’. On TikTok, people comment ‘for the algorithm’, both to support the creators they like and to influence the type of content presented to them in their feeds. Creators on various social platforms often post requests for followers to comment and like so they will keep seeing their content in social feeds. 

Marketers and advertisers need to wade into the battle for attention, at the algorithm level.

Incrementality mindset required

Currently, machine learning models for marketing campaigns are trained to pick up near-conversion events. These models result in campaigns that target people with the highest propensity to buy the brand or buy the product at that moment in time. What’s wrong with that? 

AI models, in programmatic in particular, are very good at finding those people that would have bought the brand anyway and claiming conversion credit for everyone that clicked on the ad and bought the product. Thus, campaigns seem like they’ve performed better than they actually have because results are artificially inflated. Compounding this issue is an obsession with last click attribution. 

But marketers don’t just want campaigns that look like they have performed well. They want more bang for the buck and more bottom line results than that! At least their companies’ boards and stakeholders and shareholders do. 

By 2023 we will have trained the machines to effectively measure for incrementality. This is the measurement of the impact of a single variable on an individual’s behaviour. In relation to advertising, it’s been described as the measure of the lift that advertising spend provides to the conversion rate.  

What we’re saying is, machine learning will be trained to identify the people that would buy the brand or product regardless of whether or not they saw an ad, so marketers don’t waste money advertising to them. 

We’ve already begun building the models required for incrementality, but first, there are obstacles that must be overcome. Marketers, advertisers and platforms need to move away from training to conversion goals and get over an obsession with last click attribution. Then there will be an appetite for change. 

The move away from last click towards incrementality will progress through multi-touch attribution. Our industry’s use of technology will progress to the stage where mature users of incrementality will focus their advertising budgets on people who need more effort to convert.

Focus for 2021: Marketers should expect more from marketing and advertising data. Don’t accept the same old same old measurements, targeting approaches and conversion attribution. As you, your partners and preferred platforms increase the use of AI, make sure you’re exploring how to better train the machine learning models to increase your advertising reach and performance to those who wouldn’t have already bought you anyway.

How random

How do we break into people’s bubbles and reach more new people when AI has them tightly defined based upon previous purchases and consumption? How about a zero targeting approach? Pure randomisation. Then, whoever responds to the ad, responds to it. That becomes the seed audience, and you start to build your understanding of campaign effectiveness and of the target audiences from there.  

You’ve got to get your cost of randomisation right. Every time you do segmentation of the audience you’re increasing your CPM. So you have to balance between No Targeting at $0.50 CPM versus Targeting at $15 CPM and work out what’s the best way forward. 

For one of our client’s 2021 programs, we’re not relying on existing data to design their next campaign. We are effectively going to use the response to a direct mail tactic, to product test and build the next logical product model. We’re going to shotgun it out there, measure results and use that to analyse and iterate. 

Focus for 2021: It’s a good time to embrace a return to data planning, direct marketing and a resurgence of creative testing, really, following the asterisk year that is 2020. You can’t rely on the data captured in 2020 to be a natural progression from 2019, or to help anticipate 2021. And you might as well start getting used to a post-cookie world, before it is forced upon you. 

Anticipating an attention recession

Between now and 2030, the commodity that is human attention could be increasingly fragmented, ever-more tightly protected, and more intensely subjected to the influence of algorithms that create bubbles based on likes and preferences. We can imagine whole new communities of extreme off-gridders and anti-social-networkers, as digital natives tightly embrace AI and connectivity (and who knows what else to come) and then swing back the other way.   

By 2030, marketers and advertisers could be competing with even more information and noise, in an all-out war for attention, amidst The Great Attention Recession. What are you going to start doing about it, in 2021?

James McDonald and Tom Evans are the co-founders of independent, fast-growing, full-service media agency Audience Group and Ron Ramaiya heads up Audience Analytics.

Photo by Leyy . on Unsplash.


You Might also Like

Leave a Comment