23 Sep A Guide to Ad Targeting with Artificial Intelligence
Facebook had been successfully running its ad platform for several years before launching Lookalike Audiences in 2013 to enhance its ad targeting. Since then, Facebook has grown their quarterly revenue from ads from $1.5B in the first half of 2013, to $9B+ in Q2 of 2017. Amazingly, the Facebook and Google duopoly currently own 60% of the revenue in the US ad market, and growing. No other ad platform in the US has more than 5%.
And marketers and advertisers love Facebook ads. Audience targeting is accurate simply because Facebook knows so much about their users. And these users are online all the time, to the tune of 1.3B daily active users. Better targeting = better ROI.
So how can businesses squeeze more ROI out of Facebook ads? At $9B in ad spend per quarter, a little more ROI for businesses is a lot of money in the bank. Let’s take a look at how Artificial Intelligence can lend a hand.
Utilizing Predictive Product Recommendations
First, let’s get this out of the way: retargeting absolutely is not the same as predictive targeting. Companies that help you follow users around with products they have seen already (retargeting) can get you quick and easy attribution. But how many of those users would’ve bought anyway without the retargeting? Predictive targeting takes a lot more into account than just a user’s browsing history, and will often uncover hidden patterns that are sometimes counter-intuitive. Not to mention the entire ad retargeting model is in danger.
Companies that have true predictive product recommendations will utilize a greater variety of user and item affinities, and these recommendations will learn over time from thousands of signals, and keep changing and adapting along with your users.
So how can this help with ads?
At ReSci, we took our true predictive product recommendations, and we… flipped them. Now, not only do we know which products each customer of your business really wants, but we also know for each product, on any given day: who are all the users that have an “affinity”. And these lists change every day, along with each user’s activity.
So now, you can grab a a list of the best users for each of your products, and run a Lookalike Audience to find more people who are really into… maybe this?
Hint: ReSci knows exactly which users want this Pink Fur Storage Cube
Only a platform that has real predictive recommendations, that adapt to each and every user, can bet their algorithms to use ad targeting in this way. And if you wanted to target these specific users whenever they visit Facebook, you can just run a direct ad with the product recommendation as well.
And for ReSci’s clients, it actually works.
Finding more VIPs
By now, everyone has some rough RFM models to determine who their best and worst customers are. But to build a system that truly adapts, and takes into account activity and demographics is another beast.
ReSci has a predicted highest customer future value (CFV) segment, which is different than customer lifetime value (CLV). CLV is a self-fulfilling prophecy, as it just models out from lifetime spend. CFV tries to predict how much a user will spend in the future. So if you see someone with high CLV but low CFV, look out! One of your best customers is starting to slip away.
In our opinion, a VIP is someone with high CFV, as you’re looking for your customers who are going to keep spending with you. Not just someone who has spent a lot in the past.
So just like in the product recommendation example above, run a Lookalike Audience on your ReSci highest CFV list of users, and find more (real) VIPs. Or target your VIPs with a direct ad that rewards their VIP status with a special promotion or giveaway. Give this a shot, and ad targeting will never be the same.
Alternatively, you can use our churn model, which crunches just as much data and learns as frequently as all our other models, to run exclusion lists on your Facebook ads. Boooo churners!
We also have automated audience syncing with Facebook, so getting Lookalike Audiences powered with ReSci’s predicted segments is a button click away.
The future of marketing lies in novel ways to use Artificial Intelligence to complement your campaigns and processes to boost your engagement rates and revenue. We hope you’re as excited as we are for what’s coming as the innovators keep innovating in this space.
We cannot solve a problem by using the same kind of thinking we used when we created them. – Albert Einstein
About the Author
Derek Kwan is COO and VP of Product for Retention Science. With 15+ years experience in marketing and ad tech, Derek previously led product innovation at Yahoo!. He also trains his game theory skills in his spare time as a poker player. Follow Derek on Linkedin!
Check Out Our Free AI Guides: