Audiences
Overall
Innkeepr's audiences use machine learning to identify relevant users into custom audiences across various channels. To identify those patterns we leverage customer behaviour from your Shop or Web app and translate those data into machine learning models that compose our seed, retargeting and exclusion audiences.
Types
Types are a way to configure audiences for different use cases. We currently support types for three different scenarios: acquisition, retargeting and spend protection.
Since each scenario requires audiences to be composed in different ways, we crafted each audience type is composed and updated in a way that ensures best campaign results. The corresponding audience types are:
Seed
Retargeting
Exclusion
Here is a breakdown of what you need to know:
Seeds
Seeds are a precise picture of how your most relevant users look alike (e.g. CLTV or purchase probability). You can use them as audience signals in a Performance Max campaign or build lookalikes from on Meta or TikTok.
Retargeting
Retargeting audiences are meant to target specific users who have previously interacted with your brand or website. You can add specific filters like new customers only or add to cart abondoners to be as precise as possible.
Exclusion
Exclusion audiences work as inverted retargeting audiences - think about a retargeting list with 50% of your worst customers / visitors. Those audience you will use as an exclusion list (similar to existing customer lists) to exclude irrelevant traffic from your campaigns.
Goals
Goals allow to optimize audiences for different steps in your marketing funnel. By specifying an audience goal like "checkout_completed" or "login" you can train our AI to compose audiences with users that have a high likelihood of reaching the conversion once exposed to your marketing signals.
There are different ways to set up goals in Innkeepr:
via standard integrations like Shopify,
or can by tracking user events with Innkeepr.js.
Treatments
Understanding Treatments
Treatments are a way to tailor audiences to specific campaigns. They are equivalent to ad sets, asset groups, or ad groups that live in your Google, Meta and TikTok accounts.
Once you connect a treatment to an audience, Innkeepr's Audience Optimization engine starts learning from daily traffic derived from those adsets, ultimately tailoring the audiences to users most likely to be relevant to the connected treatments.
Using single vs. multiple campaigns with one audience
While treatments are a great tool for optimizing campaign targeting, it's crucial for optimal performance to strike the right balance between your audiences and the treatments they learn from. A good rule of thumb is that ad sets or campaigns connected with an Innkeepr audience should generate at least 100 conversions per month. If this mark can not be achieved through a single campaign it could be worth connecting multiple campaigns.
Profile & Model Updates
Profile updates
Innkeepr sends targeting events at different intervals based on audience type. For seed audiences, events are emitted every three days to ensure regular updates on the most relevant users for your campaigns. For retargeting and exclusion audiences, events are sent daily to keep your lists current and effective for personalized targeting.
Model updates
The AI models that build your audiences are updated every two weeks. These updates incorporate the latest data collected via Innkeepr.js, ensuring your targeting remains accurate and reflective of user behavior.
Last updated