All articles Network & IP

Website Visitor Opt Out: Run a Public Suppression List

Translucent record card lifting away from a glowing grid of connected nodes in a dark room

Every team that identifies companies on its website eventually gets the same message: please stop identifying us. A website visitor opt out is only as good as what happens after that message. If the request lands in one person’s inbox and stops there, the company shows up again on the next report and the promise you published is false.

A public opt-out is the entry point: a page anyone can reach without an account, a form that takes a domain or an email address, and a plain statement of what changes. The suppression list is the enforcement: the durable record your identification pipeline reads before it names a company, scores intent, or writes anything to your CRM.

Below is how we handle it in Prynt Reveal, and how to keep one list consistent when the same visit also touches a CRM, a warehouse, Slack and an outreach tool.

Separate the three requests that look alike

Opting out of identification

This is the one Reveal owns. A company asks not to be named when it browses your site. Nothing else changes: the session still counts, but as anonymous traffic with no company record, no intent score and no CRM delivery. Keep the scope explicit — the exact domain, whether subdomains are included, and whether the request covers people or only the company.

Opting out of tracking, and statutory requests

A cookie-level objection and a statutory do-not-sell-or-share request are different jobs with different owners, even when the outcome overlaps. Log them separately. The first is a consent boundary your analytics must honor; Prynt Analytics also honors Global Privacy Control, which arrives as a browser signal rather than a form submission. The second has response deadlines and language set by law, so confirm those with counsel instead of guessing them. The working mechanics of our own process are described in the Reveal privacy notice: https://pryntid.com/legal/reveal-privacy-notice

Write the promise in one screen

No login, no “contact sales” step. State the scope (domain, subdomains, person), the effect (no identification, no scoring, no enrichment, no outreach drafts), what stays (aggregate visit counts), and how confirmation is sent. The page itself should be the form: https://pryntid.com/legal/reveal-opt-out

Choose a key that survives imports

Domain is the practical key for company-level suppression, since people are linked only from first-party data. Normalize before storing: lowercase, drop the scheme, drop www, drop paths and query strings. Keep one canonical row per key so two entries for the same company cannot drift apart.

Rows of frosted glass panels on black, with one entry dropping out at the second panel

Decide once whether sub.example.com matches a suppression on example.com, and encode that decision in the lookup rather than in a person’s memory. Store the minimum that makes the record auditable:

key: example.com
key_type: domain          # domain | email | person
scope: company
reason: opt_out_request
source: public_form
created_at: 2026-11-12T14:02:11Z
reviewed_by: risk-ops

For person-level requests, key on the exact email address the person submitted, never on a guess about who they are.

Enforce before enrichment, not after

The position of the check decides whether it means anything. If suppression lives in a dashboard filter, the record already traveled through enrichment, scoring, the CRM and the Slack ping — you honored the request in the report and ignored it in the pipeline.

Make the order explicit:

  1. Raw signal arrives: request, network, company candidate.
  2. Suppression lookup. On a hit, drop the company and keep anonymous counters only.
  3. Enrichment.
  4. Intent scoring and ICP fit.
  5. Playbooks, AI-drafted outreach, CRM delivery, Slack notifications.

Step two should return a decision your code acts on, not a filtered list. Also handle suppression that arrives mid-quarter: an opt-out should remove the account from active segments, cancel queued drafts and flag the existing CRM record, not just the next visit.

Propagate one list to every tool

One source of truth, and downstream tools only receive. Three workable patterns: check the list at decision time through an API (freshest, no sync gap), push a webhook on every change, or run a scheduled export for the warehouse. Pick one primary path and treat the others as mirrors.

Glowing network node sending thin blue lines to hub nodes, one hub muted and disconnected

When you add an integration, add it to the propagation checklist and test it with a canary domain you control. Include imports: a CRM import that reintroduces an opted-out domain should be caught at load, not after a sales rep sends a note. If an external service touches the data, list it — our subprocessors are published at https://pryntid.com/legal/reveal-dpa — and keep the list current so a procurement review does not stall.

Keep retention and review boring

Keep the smallest record that proves both the request and the enforcement: the key, the date, the method, who reviewed it. Decide a review cadence and write it down, because suppression is not something you want to rediscover during an audit.

Deletion requests need their own path. A common approach is to remove the identifying data and keep a keyed, non-linking record so the same company does not reappear on the next import — but the right answer depends on your jurisdiction and your contracts, so have counsel sign off rather than copying a pattern from a blog post.

Finally, test the flow like a stranger would. Submit from a fresh browser with no account, confirm the acknowledgement, check the suppression row and its timestamp, then browse again and verify that no company record, no score and no CRM entry appear. Log the result each month; a failed step is an incident, not a footnote.

FAQ

Person in a dark studio holding a translucent pane lit by cool blue light

Does an opt-out delete data already collected?

Not automatically, and the two flows should stay separate. An identification opt-out stops future naming, removes the company from active segments and playbooks, and cancels pending drafts. Whether past records must be erased depends on the request type and your jurisdiction, so answer deletion requests through their own documented process.

What should the opt-out page ask for?

For company-level suppression, the domain is enough; add an optional work email if you want to send a confirmation, plus a free-text note for context. Every extra required field reduces completion, and a form that is hard to finish undermines the promise it represents.

How fast should a suppression take effect?

Before the next identification decision. That is immediate if the lookup happens at decision time, and up to the next run if the record only lives in a downstream tool. Test the path end to end instead of trusting a setting.

Run a public opt-out you can prove. The opt-out page and suppression list are part of Prynt Reveal — start at https://pryntid.com/legal/reveal-opt-out, and see how the rest of the suite keeps the same record honest at https://pryntid.com.

Try it free

Prynt is device intelligence with a free plan: 5,000 identifications a month, visitor IDs with confidence and bot detection — no credit card. Upgrade to Pro for VPN, proxy, tampering and behavioral signals.

Keep reading