What is identity resolution?

The short definition

Every customer leaves scattered traces. They visit a site on a phone, open an email on a laptop, and buy in a store. To a marketing system those can look like three strangers. Identity resolution tries to work out which traces belong together, then links them in a single profile.

The term covers two related jobs. Some companies use it to tidy up data they already hold, such as merging duplicate customer records. Others use it to put a possible name on someone who has never identified themselves, which is usually called website visitor identification. Leafprint does the second job, for cannabis businesses.

How identity resolution works

Most systems follow the same broad pattern. They collect identifiers, compare those with a reference set known as an identity graph, decide whether there is a match, and then update a profile.

• Collect. A website pixel or another data source gathers signals such as IP address, device, browser, and on-site behavior.
• Compare. Those signals are checked against information held by the business or by third-party data providers.
• Decide. Rules or models determine whether the signals likely belong to a known person or company.
• Connect. A match is added to a profile, where it can be reviewed or sent to other systems.

Deterministic and probabilistic matching

Deterministic matching links records through an exact shared identifier, such as a matching email address. Probabilistic matching uses statistical likelihood across signals like device and location. In general, deterministic links are more certain but cover fewer people, while probabilistic links reach further with more uncertainty. Our guide to both methods goes further.

Person-level and company-level identification

On a website, a match can point to an individual or to the business behind a visit. Person-level results matter most to consumer retail, while company-level results suit anyone selling to other businesses. See how the two compare.

What identity resolution is used for

• Understanding traffic. Seeing who your campaigns attract, not just how many visits arrive.
• Judging channels. Comparing sources by the people and companies they deliver.
• Finding accounts. Spotting businesses that are researching what you sell.
• Cleaning data. Connecting duplicate records across your systems.

Limits you should know about

No system identifies everyone. Results depend on the audience, the geography, and what data sources hold, and privacy tools can reduce them further. Some matches will be incomplete or wrong, so vendor match rate claims deserve testing on your own traffic. Our vendor evaluation guide explains how.

Privacy and consent

Connecting behavior to people raises real privacy questions. Businesses should disclose the practice, offer a way to opt out or request deletion, and review applicable state laws with counsel. A match also does not create permission to contact someone. In particular, a matched phone number is not SMS consent.

Identity resolution in cannabis

Cannabis businesses have narrower advertising options than most industries, which makes each earned website visit more valuable. Many shoppers browse a menu and never join a loyalty program, so identity resolution can show a retailer who that interest may be. Cannabis also raises the stakes on discretion and compliance, so the data needs careful handling. Leafprint is built for this market, and you can see how it works or read how dispensaries use it.