Types of Fan Linking

The Fan Data Platform supports two types of fan profile linking, the first is known as Hard Linking. Hard Linking is set up to follow a strict set of rules where certain criteria must be met for two fan profiles to be linked together. The second type of fan linking is known as Probability Linking. This method uses a FanThreeSixty proprietary AI algorithm to match similar profiles. The algorithm creates comparisons of fan profiles using the fan’s name, email address, home address, and other demographic data to generate a match probability of those two fan profiles. The goal of this model is to find fan profiles that are not exact matches (which our hard linking criteria should pick up on) but are still similar enough to accurately link together and reduce duplicate fan profiles in our clients' fan databases.

Hard Link Example

Link Status: Hard Linked ✅

This method of hard linking is highly efficient at linking fan profiles that we know are the same people based on the strict requirements. There are other fans that can and should be linked on a slightly looser set of requirements. These would include similar names, emails, phone numbers, and addresses/postal codes.

Source-Specific Hard Linking Logic

Data Imports use a special set of rules to allow for fan linking. Since these data sources do not always have the full breadth of demographics data, the system only requires and email or phone number to match in order to link these records together.

Probability Link Example

Link Status: Hard Linked ❌

Link Status: Auto Linked ✅ (Overall Match Probability - 99.9%)

Each data point gets compared and similarity score is generated. Once all fields are scored, an overall match probability is then generated and used to determine whether or not the profiles should be linked.

At this time, we will be setting all organizations to auto link fan profiles which have a match probability of 90% or more unless you have discussed a different match threshold with your Client Success representative.

Data Enhancements

In addition to data gathered from client systems directly, FanThreeSixty also ingests third-party data that allows us to update outdated information and fill in gaps on fan profiles. This added data will allow for more profiles to be matched (based on a historical analysis, about a 20% increase in potential links) and ultimately this reduces a client’s fan profiles by 2% by linking duplicate profiles upon the initial run of our enhanced linking model.