Skip to main content

AI AudienceTM

Optimize Ad Spend with Laser-Focused Targeting

Boost campaign ROAS by 2-3x the original campaign.

Trusted by global leading brands and enterprises

Triple Your ROAS. No Extra Spend.

By predicting your customer intentions, AI Audience displays your ads to the right visitors that will purchase – boosting results by at least 2-3x the original campaign ROAS.

Outsmart Ad Platforms

Remove the guesswork – save your campaign budgets from being wasted on inefficient targeting and scale your spend to achieve higher profitability.

Tailor-Made Machine Learning Model

Our model drills down to the most detailed level so you can get the most value out of your first-party data.

Continuous Data & Model Updates

Always target the most engaged users with weekly audience data updates.

Link Your Store with Major Marketing Channels

Initiate and expand your advertising endeavors across Meta, Google, TikTok, and others

Featured Case

Advanblack achieves 300%+ campaign ROAS using AI Audience™ for retargeting

 

Read more

+ 389%

increase in ROAS

– 23.7%

decrease in cost per purchase

We are so glad to collaborate with MoBagel. By utilizing AI Audience, the team created a model that converted our website visitors to customers. We saw a 228% increase in ROAS for our retargeting campaigns.

Global One ClickAgency Partner

Optimize your campaign in just 4 steps

We’ll walk you through the setup at no cost.

Step 1
GA4 Audit
Data Reporting
Step 2
AI Audience
Model Training
Step 3
Campaign
A/B Testing
Step 4
Performance
Review

We are excited about partnering with agencies to deliver value to more brands.

Get started with a free trial

Only pay when you get results

FAQ

Is AI Audience GDPR compliant?

Yes, we adhere to GDPR and data privacy policies. PIIs are not stored within our platform, and we use first-party distinct identifiers across platforms to enable secure and compliant communication. User-level data are directly sent in real-time to Facebook’s server via a ‘passthrough’ method for customer matching, meaning we do not store this data.

Do you share data with third parties or combine data with other clients?

No. Our client’s data are only used to train machine learning models and to deploy customer predictions for the same client. We also follow internal information security processes that are compliant with ISO 27001 standards.

What ad networks does AI Audience support?

Currently we support Meta Ads and Google Ads.

How does your AI audience differentiate from in-platform lookalike audiences?

AI Audience segments users that have previously interacted with a brand, whereas in-platform lookalike audiences finds new audiences based on an audience signal. The two can work synergistically as we have found great success with AI Audience in combination with Meta’s Advantage+ Lookalike and Google PMAX campaigns

How is cookie-ing users more accurate than sending PII via conversions API?

Using PII improves audience match quality but not coverage. Since PII is not available for majority of website visitors, relying on PII means that only a small subset of visitors can be matched and retargeted. On the other hand, first-party cookies can be matched to ~90% of website visitors through Facebook’s CAPI’s Advanced Matching. That being said, AI Audience is not restricted to only cookie-level data. Depending on a client’s website and privacy policy, we are able to hash PII to send along with cookie-level data to improve match quality.

How is your audience more effective than breaking out retargeting audiences by number of pages viewed or time on site?

Our model consistently performs 1.5X-3X better in recall and precision when compared to simple heuristic-based models such as page views and engaged time. This is because our machine learning models are based off of 300+ carefully selected features to analyze and predict user behavior, allowing us to more accurately capture customer intent.

What is an ideal audience size for AI audience?

In the world of ad audiences – size matters. To determine whether or not an audience is ready, we look at its size. From our experience, we generally aim for a target audience size of ~10K to run an effective ad campaign.

If audience size is too low, the ad network isn’t able to reach enough people which would result in the same people getting the same ad multiple times (not good); and if the audience size is too high, it means that the campaign is targeting too broad of an audience, which defeats the purpose of segmenting audiences via AI.

There are a couple ways to estimate audience sizes:

  1. Using the AIA_Tag event (the best way)
  2. Count fbp (not as great way)