AI scale micro targeting is coming fast

A webpage with one of Simon's videos being recommended to them with Simon thinking "this feels very familiar"

Targeted online ads have so far promised a lot more than they delivered.

Having worked with targeted ads for many years, my experience is that they work more by accident than design. If you throw enough ads at enough people some of them will convert by default.

However, the world of targeted advertising is about to change – drastically!

Thanks to AI, ads will go beyond ‘targeted’. The big advertising platforms like Meta, and even Mailchimp, are working on AI-generated marketing messages and content that is created and tailored for individuals.

Targeted ads are not so targeted

Currently the large ads platforms place people in ‘cohorts’ or ‘groups’ that are based around interests or life stages. This grouping approach allows marketers to target, say, parents who are buying a house, players of online games, fans of horse racing, and so on.

Such targeting feels specific, but it is often much broader than it appears. I see these platforms bombarding vast numbers of people with ads in the hope that someone will convert and deliver the results they’ve promised.

AI is about to narrow the focus – to one.

Before the advent of transformer models – a type of neural network developed by Google and what enabled OpenAI to launch ChatGPT so fast – it simply wasn’t economical or even effective to process everything you knew about a user and then categorise them according to interests and life events. There was just too much data to process and the available tech wasn’t up to doing the sorting and labelling.

With advances in AI, mass data processing is now much quicker, easier, and more effective.

What does this mean for advertising and marketing?

AI means much more fine-grained targeting for ads. Instead of grouping users based on a few rudimentary things they have in common, you will be able to target people according to very specific behaviours and signals. (This might sound like a good thing, but bear with me…)

It will soon be possible to take all the things you know about a user, including their location, browsing and purchase history, and build a very clear model of what that user likes and dislikes, and where they are in their life – all to a very, very detailed extent.

All this information can then be used to create content just for that specific user profile.

Google, Meta and the many data brokers who track every moment you spend on your phone and computer have this data ready to go. Some of it goes back years or even decades.

The tools these platfroms are developing will make a video advert, based on things that user likes and responds to, and show them that as an ad on Instagram or TikTok. Mailchimp will be able to write an email created for each user based on what it knows they will respond to in terms of sales offers. All of this content will use location data to feature places a user likes or has visited. The people who appear in this content will talk and act in a way that the AI knows they will respond to. All of it will feel spookily familiar and personal.

What’s more, search results will be increasingly tailored and written in a way that Google knows is most likely to match a user’s search goals. If someone is looking for a recommendation, it will show them sponsored content in the way that is most likely to lead to a conversion – an instant blog post that explains the search object in detail, a video that shows exactly how something works, or maybe just a link straight to the checkout if it knows the user’s in a rush!

The speed at which content can be created on demand by AI is almost unbelievably fast. The amount of data that the big advertising companies have collected on users (on you!) is vast and grows every time we use our phones or computers.

Is this even legal?

I could do a whole blog on this (and may well do) but the short answer is: it remains to be seen. Privacy laws aren’t clear on the use of offshore data sets that are fed into the AI models that are then used in the UK or EU. Because of the way AI ingests and processes data it’s impossible to say if it has used your data to generate it’s output, or even how it uses it. So, if a model is trained using data from a data broker based outside of the privacy law jurisdictions it would be impossible to prove whether or how that data was used.

The challenge: stand out from the AI crowd

Let me be clear: I am not in favour of this.

I think it will be a disaster for many of the platforms that use AI micro-targeting and will turn many online spaces into sponsored content hell-hole.

So why am I telling you this? I’m not trying to sell you a space on my device and advert-free island paradise; although thats increasingly looking like a good business idea!

What I want to try and get across here is that when everything is micro-targeted, you need a way to get your business noticed. Showing up as you, for the people who want to learn more about you and the work you do, is increasingly important. It will also be increasingly rare – which is how you get noticed.

Building your own audience that you have a direct connection with via a podcast and email list is the most effective way to stand out from the crowd and more and more people are doing it.

The big platforms have to be greedy and invest in everything AI just to keep their share price high. It’s going to get messy and I believe these platforms will go the way of Myspace and AOL and become relics of the past.

Gathering your own group of interested, curious and engaged humans will become the way that marketing works well for those who are “selling” connection, understanding and transformation.

But… AI could replace me altogether, how do I avoid that?

Everything is changing and that’s not going to be easy. For coaches and consultants offering a generic service with price as the focus, AI will swamp their marketing, and maybe them altogether

However, if you have crafted a unique story and your offer is built around your own method or process, you have what it takes to cut through the AI-generated advertising.

Building your own audience now, with email list and podcast, is a great way to start reaching those who want to hear a human voice.

I believe people want to find people who are not replicants; people who are offering connection and interaction on a human level. Showing up and offering just that will be a breath of fresh air in an AI-dominated marketplace.

If you want to connect with me, another human, then you can listen to my podcast and become a subscriber below to receive some Monday marketing motivation.