How identity resolution connects anonymous visits to B2B pipeline

October 18, 2026

Identity resolution B2B attribution header: three broken anonymous visit trails bend into one bright KNOWN node, then a single solid buyer path climbs to the upper right

Identity resolution is the step that connects anonymous website visits to the real person, and then the account, behind a deal. Without it, every visit before the first form fill stays unattributed, and the channel that started the buying process gets no credit. In B2B marketing attribution, it decides whether your report begins at the first ad click or at the demo request.

Most buyers visit several times before they tell you who they are. They read a blog post, come back from a LinkedIn ad, check pricing on their phone, and only then sign up. Identity resolution is how those visits end up on one path instead of looking like four strangers.

What does identity resolution mean in marketing attribution?

A tracking pixel gives every new browser an anonymous visitor ID, usually stored in a first-party cookie. Each page view, referrer, and UTM tag is recorded against that ID. At this point you know a lot about the visit and nothing about the visitor.

Resolution happens when that visitor gives you something that identifies them: a work email on a signup, a demo form, a webinar registration, or a click from a tracked email. The visitor ID is matched to that email, the contact is created or found, and the earlier anonymous visits are attached to it. From then on, the contact's history starts at the first visit, not at the form.

There are two ways to match. Deterministic matching uses a hard identifier, such as an email address or a login, so the match is exact. Probabilistic matching infers a match from signals like IP address and device type. That is weaker in B2B, where many people share an office network or a VPN. For attribution, rely on deterministic matches and treat probabilistic ones as hints.

How does an anonymous visit get tied to a real person?

Take a common path. Visit one is a LinkedIn ad click. Visit two is a blog post the buyer found through ChatGPT. Visit three is the pricing page. None of these carry an email, so all three sit under the same anonymous ID. When that person signs up with a work email, the three visits join their contact record, and the meeting and deal that follow land on the same record.

If that join never happens, the report shows a contact whose story starts at the signup, often labeled Direct or organic. That is one of the main reasons Direct keeps growing in B2B reports, which we covered in B2B attribution myths that inflate Direct traffic. It also explains why first-touch reports drift. First-touch attribution only works if the first touch is still attached to the person.

Forms matter here too. If your form or CRM overwrites the original UTM values on submit, the join happens but the source is lost. Our list of UTM mistakes that send B2B deals to Direct covers the hidden-field setup that keeps them intact.

What happens when the same buyer switches devices?

Cross-device journeys are where identity resolution gets hard. A buyer clicks an ad on their phone during a commute and signs up from a laptop at work. Those are two browsers with two anonymous IDs. The laptop gets identified at signup. The phone stays anonymous unless the same person identifies there later, for example by logging in or clicking a link in an email you sent.

Once both browsers are tied to the same email, the two histories can merge and the phone ad click returns to the path. Until then it is missing. That gap is normal, and the fix is more chances for a deterministic match: tracked links in onboarding emails, logins on mobile, and a "how did you hear about us" field that catches what the pixel cannot.

Browsers also limit how long an anonymous ID survives. Safari's Intelligent Tracking Prevention caps cookies set by JavaScript at seven days, so a buyer who waits two weeks between visits can look brand new. Consent banners add another gap when visitors decline tracking, and how cookie consent changes marketing attribution covers what you can still measure then.

Should you resolve to the person or the account?

Both. B2B deals are bought by groups. A champion finds you through a podcast, an analyst reads the pricing page twice, and a director books the demo. Each one is a separate person path. If you only credit the person who booked the demo, the podcast that started the deal disappears.

Identity resolution at the account level: champion, analyst, and buyer person paths at the same company domain roll up into one account path from podcast to pricing to demo to deal — Person paths roll up by company domain, so the account path keeps the first touch from anyone on the buying team.

Account-level resolution groups those people by company, usually by email domain or by the account record in a CRM like HubSpot or Salesforce. The account path then shows the first touch from anyone at the company and every channel along the way. Our guide to ABM attribution without losing the buyer's path goes further on account-level reporting.

Personal emails break the domain rollup. A signup from a personal address will not group with the company until someone associates the contact with the right account. Duplicate contacts and accounts cause the same kind of split, which is why CRM hygiene that keeps B2B attribution trustworthy belongs in any identity conversation.

How do you check whether identity resolution is working?

Pull a sample of recent closed-won deals and open each contact's path. Four checks will tell you most of what you need to know.

Identity resolution checklist for B2B attribution: visits before the first form, Direct falling for known deals, UTMs kept in hidden fields, and contacts tied to accounts — Run these checks on recent won deals before you change attribution models.

Visits before the first form. If nearly every won deal starts at the form fill with no earlier visits, anonymous visits are not being joined.
Direct falling for known deals. After you close identity gaps, the share of won deals credited to Direct should drop. If it does not move, look at consent, redirects, and form fields.
UTMs kept in hidden fields. Submit a test form from a tagged link and confirm the original source survives in the CRM.
Contacts tied to accounts. Look for won deals with a single contact on the account. The rest of the buying group often exists as orphan contacts.

These checks take about an hour. If the paths are broken, switching from first-touch to multi-touch only spreads credit across an incomplete chain, so fix identity first.

Where do AI answer engines fit in?

Visits from ChatGPT, Claude, Perplexity, and Gemini follow the same rules. The visit arrives with a referrer, or sometimes without one, and stays anonymous until the buyer identifies. When identity resolution works, an LLM-referred first visit stays on the path all the way to the deal. Our guide on how to attribute ChatGPT and other LLM traffic explains how to catch those referrals in the first place.

How Source handles identity resolution

Source installs with one pixel or Google Tag Manager, with no data warehouse to build. Drop in the pixel and Source stitches visits, page views, meetings, and closed deals across the tools you already run, so the anonymous visits before a signup stay on the buyer's path.

On that path you get multi-touch, view-through, offline, and LLM attribution, plus real CAC, goals, and channel quality. You can chat with your data, get AI recommendations, and receive daily Slack reports. Source is SOC 2 Type II and ISO 27001 certified, and customer data is not used to train external models.

Start free at dash.source.app/signup and see where your buyers' paths really begin.