You can’t get any clearer. “Identity” means knowing who your customers really are. If you’re a marketer who understands the importance of putting customers first, it’s crucial to get identity right.

Identity resolution fuels a data-driven marketer’s mission to build genuine and lasting connections with people who truly want to buy their product or service. The connections need to happen at specific points along the buyer’s journey, and in the moments that matter most to the customer.


Two types of identity signals

You need a complete and accurate picture of your customer to start this process, and this is where identity makes its key contribution. There’s a growing number of identifiers attached to each touchpoint in a buyer’s journey. These identifiers are the data marketers use to figure out who’s who and where they are on the buyer’s journey.

There are two kinds of signals data scientists look at when establishing and validating a customer’s identity.

  1. Deterministic
  2. Probabilistic

Deterministic signals are identifiers that can be explicitly connected to a customer, such as an email or username a person uses to log in to your company’s website, and the cookie that links to it. On the other hand, probabilistic matching uses pattern recognition to calculate likely connections between customer signals and devices. Examples include IP addresses, browser or device type, operating system or location. These identifiers infer a relationship across identity signals that help to build a statistically likely customer picture.


Is one better?

Probabilistic matching allows for greater scale, but it can be inconsistent when it comes to accuracy. It’s why some marketers prefer using deterministic data because it’s more definitive. It requires identifiers explicitly provided only by the customer.

It turns out though, that up to half of deterministic data can be incorrect. Or it can be fraudulently created by bots. Here’s a simple example of how this happens. Let’s say your overly generous son or daughter shares your Netflix login information with a handful of friends. There’s now a group of devices attached to your email address – and they’re reporting viewing habits that likely don’t reflect your own.

Here’s another scenario. You order from an online store and have it shipped to a friend’s house because you live in an apartment that makes package delivery a hassle. Your name is now attached to an address that’s not your own.


How artificial intelligence can help

This is why many marketers believe probabilistic data turns out to be a better approach. It looks at a wide variety of data, so it can be used to eliminate false deterministic signals. These multiple identifiers can be used by machine learning and artificial intelligence to determine links with high confidence.

Removing false deterministic signals and complementing them with probabilistic signals creates an effective hybrid approach to identity – especially when both types of data is compared side by side using artificial intelligence. This powerful approach determines which data points are actually linked. The rest on both sides can be ignored. This process is known as identity resolution. It provides the most complete and persistent view of people who want to buy your product or service.

It allows you to have high confidence that you have identified who your customer is, so you can reach out and connect with him or her in a privacy friendly manner at the moments that matter most along their buyer’s journey. Plus, you also create the ability to scale your marketing to others like them.

Schedule a call to learn how we use deterministic and probabilistic data signals to identify the top 3 percent of people who are actively looking for your product or service in your market.