The digital space has been abuzz recently with talk of Machine Learning and Artificial Intelligence. You’ve probably heard it mentioned in the news – some aspects are intriguing, some are eye-opening and some are flat-out mind-blowing.
Artificial Intelligence, or A.I., and Machine Learning are terms we use to describe how online advertising networks (namely Google and Facebook) define and target market segments.
The two terms mean different things, so they’re worth defining:
- Artificial Intelligence, or A.I., is the science of making machines smarter,
- Machine Learning is having machines learn to be smarter, themselves. It’s functionality that helps software perform a task without explicit programming or rules.
When applied to advertising A.I. and Machine Learning works to show the right ad, to the right person, at the right time and place, in order to bring about the right result (a purchase or a lead enquiry). And all of this at the most cost-effective spend for the advertiser.
Google provides real-time data by analyzing 70 million signals within 100 milliseconds!
Google has 1 billion users across their 7 properties (Google Search, Maps, Chrome, Play, Gmail, YouTube and Android). It looks at both real-time data & historical data to differentiate between interest and intent and identifies potential customers through pattern recognition, grouping them accordingly.
In this article, we’ll focus on Google’s A.I.-related tools for the Google Ads network, but the same tools apply to Facebook and other digital advertising networks too.
What is A.I. when it’s applied to advertising?
A.I. and machine learning begins with data. Lots of data. In the digital space, tech giants like Google have created the necessary tools to collect real-time data from their users. This data is anonymised (they don’t have your name against that lava lamp purchase on Amazon last month), and then applied to help deliver advertising that’s smarter than ever.
How does it work? Google works to understand people and what they’re looking to do online by looking at three key factors. These are:
Let’s look at these, one by one
Who They Are
Google has created audience definitions based on how broad chunks of the market behave. These are what are known as Affinity Audiences and In-market Audiences.
- Affinity Audiences are built based on people’s interests and habits. Originally built for large-budget TV advertisers to use to target their audiences on digital channels, Affinity Audiences place audience targeting into pre-defined buckets based on interest categories.
- In-Market Audiences are built from users who are actively researching or comparing products and services across the Google network (including partner sites like YouTube). For example, if someone is browsing travel websites or engaging with travel-related ads, they’re likely to be able to be included in an In-Market audience for travel advertising.
What does targeting these audiences look like in practice? Let’s take the example of Lizzie and Jess.
If we were to look at only demographic data for these two, we’d categorise them in the same box. They’re 25-35, female, professionals, living in Auckland. However, with some deeper insights, we can see that they’re quite different people, with different interests and lifestyles, who will respond completely differently to advertising and offers.
Google and Facebook audience data allows us to “see” these differences before we begin our advertising, and tweak our targeting accordingly.
Let’s look at an example of a travel company targeting the Auckland market with some packages for Southeast Asian destination getaways for families. Only a few years ago, their options would be to broadly target people in the same demographic, hoping they are parents and have the income to afford a family holiday. The advertiser might use an ad buy across print newspapers, or some static web banners or other media to reach that relatively broad audience.
Now with the market intelligence available on digital ad networks, that same advertiser can effectively target families who have already shown interest in cultural family holidays, who match a certain income bracket and maybe who are even browsing holiday and travel websites already.
I hope Jess and family have their passports ready.
Where They Are – The Context
Now we know the who – the audience we’re trying to reach with our advertising, we can look at the context of that audience; where they are, what devices they’re using, and when they’re using them.
Machine learning enables advertisers to optimise advertising campaigns for geographic location, types of devices (desktop, tablet, or mobile), and the time of day those devices are being used – whatever is likely to be the most effective for a campaign and its goals.
The good news is it’s simple to begin doing this. By setting up conversion tracking from the beginning of an ad campaign, and opting to use automated bidding strategies, advertisers can make full use of data and machine learning resources to target people who are more likely to respond to their advertising.
What They Want
The third piece of the machine-learning puzzle is the insight into what the consumer is actually wanting to do or buy.
We’ve discussed Affinity and In-Market audiences above. In the Google Ads platform, we can include audience targeting for both:
- Broad Affinity audiences in predefined groups Google has defined through search and browsing history,
- More defined In-market audiences made up of people who are actively searching for products related to a business’ offerings.
These audiences give advertisers insight into what customers are looking for, further helping improve campaign targeting people who are already interested in an advertiser’s offering.
Back to our travel example again. If Lizzie has been wanting to book a hiking holiday, and she’s been searching on travel and flight websites over recent months, then an adventure travel business can target her through using Affinity and In-market audiences in its campaigns.
Smarter Advertising. All of this means showing the right ad to the right person at the right place and time becomes much easier, and that this kind of smart advertising has become more accessible for businesses of all sizes.
Paid Search, Display and social media ad campaigns are now optimised using A.I. and Machine Learning, making digital advertising more cost-effective and finely targeted than ever. Does your digital strategy include making the most of the A.I and Machine Learning tools available?