Frequently Asked Questions

Methodology

Is your data modeled?

Yes. Our model is called TrueAffinity. Our model is based on pseudonymized engagement data collected from social networks between people and the brands, celebrities, shows, movies, and all other entertainment they love. 

How do you collect your data?

Our data is collected by capturing the positive pseudonymized engagements (both likes and comments) between people on social media and the content posted by over 70,000 brands, TV shows, movies, and other media and entertainment properties. We do not use overall page likes to train our model as we have proven them to be an inaccurate indicator of affinity.

What is the recency of your data?

Our modeled data is refreshed every 13 weeks.
Our segment data is refreshed every 30 days.

Do you use Natural Language Processing?

Yes. Our TrueAffinity model is created using only positive engagements between people and the brands, celebrities, shows, movies, and all other entertainment they love. This includes net neutral to positive comments on posted content, along with likes and retweets.

Where do you collect your model data?

We collect the data for our model from public brand and celebrity content posted across Facebook, Instagram, and Twitter.

How do you use social data?

We use social data to understand the affinity between people and the brands, celebrities, shows, movies, and all other entertainment they love. We track pseudonymized IDs of people as they engage with their favorite things across social networks. This data, along with data collected from over 550MM additional pseudonymized IDs around the world, is analyzed using collaborative filtering to understand the relationship and positive affinities between people and brands. This creates our Act-Alike model called TrueAffinity, that can accurately predict, if a person has declared interest in certain things, they will most likely have intent towards another set of other things. Our model has been proven to be a more accurate indicator of future purchase intent than leading past-purchase audience providers.

Why social data instead of past-purchase data?

We say “a purchase is fleeting, but love is forever”. This tongue-and-cheek sentiment attempts to channel renown marketing professor Byron Sharp and the work done by the Ehrenberg-Bass Institute for Marketing Science. Past-purchase is proving to be a lesser and lesser predictor of future purchase intent as overall brand and product loyalty continues to decline. So as a strategy, it is fleeting. New customers are vital to any growth or maintenance strategy as the “customer bucket” will continue to leak medium and heavy buyers over time. Light buyers have never been more important and identifying those most likely to switch is the key. Affinity is the best way to identify those most likely to be interested in switching or trying new products, TV shows, Movies, etc.

Programmatic

How do you create your segments?

Our segments are created when we match the brand engagements from our pseudonymized social data with partner data containing the same brand engagements. Our partner data sources are based on web and location data collected from Cookies and MAIDs.

What is the recency of your data?

Our modeled data is refreshed every 13 weeks.
Our segment data is refreshed every 30 days.

What countries do you cover?

Our Oracle Data Cloud segments cover North America and APAC
Our LiveRamp segments cover US
Our Eyeota segments cover US and EU

What are your top performing programmatic segments?

Entertainment (TV, Movies, Music, Gaming)
Tech/Consumer Electronics
Automotive
Beauty & Cosmetics
Politics
Restaurants
Sports

Where can I find your segments?

Our segments can be found across all major platforms and DSPs including:
Oracle Data Cloud Marketplace
LiveRamp Marketplace
Eyeota Marketplace
The Trade Desk
Adobe Ad Cloud
Amobee

Which platforms can I activate your segments from?

Our segments can be pushed to and activated from all major platforms including:
DV360
The Trade Desk
Adobe Ad Cloud
AppNexus
MediaMath
Amobee
DataXu

Which data marketplace should I use for custom segments?

Affinity Answers uses consistent thresholds for audience modeling through TrueAffinity. As we are modeling different audience pools (cookies/IDs) by data marketplace, we recommend creating custom segments across each marketplace to reach a different set of end users. Clients also find benefits of optimizing between different versions of the same segment.

Do you create Custom Audiences?

We sure do. We work with clients in a variety of ways to deliver the exact segment to meet their requirements. Here are some of the most common:

  • Create custom, composite segments using brands, entertainment properties, sports teams, celebs and more
    example: You want to target New England Patriots Fans who also love Dunkin Donuts (p.s. It’s all of them)
  • Use hashtags or common strings found in posts to create audiences around a trend
    example: You want to target environmentally conscious people
  • Use multiple branded segments to be able to accurately target a lifestyle
    example: You want to target everyone interested in luxury travel

Addressable TV

Can Affinity Answers data be used for Activating users on Addressable TV?

Yes. Affinity Answers data is available on Liveramp Connect:Select.

What are your top performing addressable segments?

Automotive
Beauty & Cosmetics
Events
Politics
Restaurants
Retail
Sports
Travel

Is Affinity Answers data PII based?

Yes. Affinity Answers data is modeled. Via Liveramp, our model (TrueAffinity®) scores partner data, which is PII based.

Since Affinity Answers data is modeled, what is Top x% in your data segments?

Our data is built for the digital targeting ecosystem, where reach is crucial. Our syndicated segments available on the shelf have the Top 30-40% of users in the segment.  

Can you create a Top 10% of users segment?

Yes. Contact us at audiences@affinityanswers.com and we can work with you to create a Top 10% segment. A segment can be ready for activation within 7 business days.