# How to Turn an ICP Into Prospecting Filters

A playbook for turning a written ICP into account filters, person filters and exclusion rules, with the filter caps and suppression features in Apollo, Sales Navigator, Ocean.io, UpLead and Clay.

A written ideal customer profile (ICP) is a description. A prospecting tool only understands fields, values and exclusions. The work in between decides whether reps get a list of real fits or the top slice of a loose search.

- Turn an ICP into prospecting filters by rewriting each trait as a testable rule on one field.
- Split rules into account filters, person filters, exclusions and manual checks.
- Run account filters first, save them as a list, then search people inside it.
- Keep permanent disqualifiers apart from suppressions such as customers and opt-outs.
- Sample the output by hand before any rep sees it.

This applies when you already have a written ICP built from closed-won and closed-lost accounts, and a prospecting tool with company and contact filters. Filters only reach attributes a vendor stores. Traits like pain, process maturity or budget stay manual checks. Sales Navigator, Apollo, Clay and UpLead also cap results or exclusions, so a broad search can be cut short.

## What It Means To Turn An ICP Into Prospecting Filters

Turning an ICP into prospecting filters means rewriting each trait of your best-fit customer as a rule a database field can match, then pointing person filters at the several roles in a buying group, which [Gartner found ranges from five to 16 people across as many as four functions](https://www.gartner.com/en/newsroom/press-releases/2025-05-07-gartner-sales-survey-finds-74-percent-of-b2b-buyer-teams-demonstrate-unhealthy-conflict-during-the-decision-process).

An ICP says "mid-sized software companies in North America that sell to finance teams." A filter says "employees 200 to 1,000, headquarters in the United States or Canada, industry equals software." The first is a description you agree on in a meeting. The second is something a database can match, count and repeat.

That translation sits at the centre of [prospect list building](https://growthtribune.com/prospecting-data/prospecting/). It produces four outputs:

- **Account filters** decide which companies qualify (industry, headcount, geography, funding, technologies).
- **Person filters** decide who inside those companies you contact (title, function, seniority, location).
- **Exclusions** remove companies and people that match the filters but must not be contacted.
- **Manual checks** cover the traits no field holds, checked by a person on a sample.

> **Exclusion rule:** A rule that removes a company or person from a prospect list even when they match every inclusion filter, such as a current customer, a competitor or an address that has opted out.

Order matters. Run account filters first and save the result as a list. Then run person filters only inside that list. If you search people directly by title, the tool returns people at companies that never passed your account rules, and you spend the review time throwing them out.

The buying-group range is why person filters need several roles. Gartner's survey of 632 B2B buyers puts the group at five to 16 people across up to four functions. A filter set built around one job title reaches one of those people and misses the rest of the decision.

## Which ICP Attributes Can Become Prospecting Filters

Headcount, geography, funding, founding year and named technologies translate cleanly into filters in tools such as UpLead and Clay, while industry needs rewriting per tool because each vendor applies its own classification, and traits like pain or budget cycle stay manual checks.

The table maps common ICP attributes to the field that carries them in the five tools this guide covers, and how far you can trust the match.

**ICP attributes, filter fields and reliability**

| ICP attribute | Filter field | Tools in this guide | Reliability |
| --- | --- | --- | --- |
| Industry | Vendor taxonomy or NAICS/SIC | All five | Rewrite per tool |
| Headcount | Employees or company size | Apollo, Sales Navigator, UpLead, Clay | Solid as a range |
| Revenue | Revenue or revenue bracket | Sales Navigator, UpLead, Clay | Estimate, weak if self-reported |
| Geography and HQ | Location or headquarters | Apollo, Sales Navigator, UpLead, Clay | Solid |
| Funding | Funding | UpLead, Clay | Solid for funded firms |
| Year founded | Year founded | UpLead | Solid |
| Technologies | Technologies or technographics | Apollo, UpLead, Clay | Varies, weak in Sales Navigator |
| Hiring and intent | Job postings, buying intent | Apollo | Signal, not proof |
| Company type | Type of business | UpLead, Clay | Solid |
| Job title | Title | Apollo, Sales Navigator, UpLead, Clay | Needs every variant |
| Function and seniority | Function, management level | Apollo, Sales Navigator, UpLead | Solid |

Three attributes need more care than the table can show.

**Industry is not one field.** Each vendor classifies companies its own way. UpLead offers NAICS and SIC codes alongside an industry category and tags. Ocean.io assigns industry categories and tags with AI, reading company websites and LinkedIn pages. An independent Sales Navigator playbook advises adding adjacent industries to account searches because companies are often misclassified. The same ICP rule, "B2B software for finance teams," therefore returns different companies in each tool. Write the industry rule once in plain words, then rewrite it in each tool's own taxonomy.

**Some filters are weak signals presented as data.** The same Sales Navigator playbook treats self-reported annual revenue, technologies used and the lead profile industry filter as unreliable. Use them to widen a search, not to disqualify accounts.

**Firmographics depend on what the record holds.** A filter can only match the [firmographics a business contact record holds](https://growthtribune.com/what-is-a-b2b-contact-database/), so check that the vendor stores the field before you write a rule around it.

Some ICP traits have no field at all:

- **Pain or trigger.** No filter states that a company struggles with manual reporting. Job postings and buying intent in Apollo act as proxies, and a person confirms the fit on a sample.
- **Process maturity.** Named technologies hint at it (a company running a specific CRM has some process), but the rule stays a manual check.
- **Budget cycle and buying timeline.** These sit outside every database here. They belong in qualification, not in the filter set.

Mark each of these as a manual check in the rule sheet so nobody mistakes a proxy for the real attribute.

## How To Turn An ICP Into Prospecting Filters, Step By Step

Build the account filter first, save the matching companies as a list and run person filters only inside that list, and split any search that reaches a tool's display cap, such as the 2,500 results LinkedIn's community guidance gives for a Sales Navigator search, by region or size band.

The eight steps below carry the decision, the input and the output of each stage. Work through them in order.

- [ ] **Write the ICP as three to seven testable rules.** Decide which traits your closed-won accounts share. Use inclusive ranges and named values ("200 to 1,000 employees", "HQ in the US or Canada"). The output is a rule sheet with one rule per line.
- [ ] **Tag every rule as account, person, exclusion or manual.** Decide where each rule runs. A rule about the company is an account filter, a rule about the contact is a person filter, a rule that removes records is an exclusion, and a rule no field holds is a manual check.
- [ ] **Map each rule to the tool's own field names.** Rewrite industry in the tool's taxonomy (NAICS codes in UpLead, AI-assigned tags in Ocean.io, industry plus adjacent industries in Sales Navigator). The output is the same rule sheet with a field name beside each rule.
- [ ] **Build the account search first and save it as a list.** Use the tool's container for companies, which is a saved company search with the list filter in Apollo, an account list in Sales Navigator, a saved search in Ocean.io, Search Lists in UpLead or a table in Clay.
- [ ] **Add disqualifiers and suppressions.** Exclude competitors, regions you cannot serve, current customers, open opportunities and opted-out contacts. Section four covers how to write each kind.
- [ ] **Run person filters inside the saved accounts.** Add every title variant ("VP Finance" OR "Vice President, Finance" OR "Head of Finance") and cover each role in the buying group, not only the economic buyer.
- [ ] **Split any search that hits a result cap.** If the count sits at a tool's limit, break the search by geography or size band and run each part separately.
- [ ] **Save and name the search.** Use a name that states the segment and the rule-sheet version, so anyone can re-run it and compare results.

Steps one and two do most of the work. A rule such as "fast-growing company" cannot be tested. "Headcount grew in the past year" can be, in tools that carry growth signals. Rewrite every vague rule until a field can match it or until you accept it as a manual check.

Step three needs care. An industry rule copied word for word from one tool to another keeps its label but changes its meaning, because the categories underneath differ.

Step six is where coverage of the buying group is won or lost. Map the roles first (economic buyer, user lead, technical evaluator, finance), then write a title set for each. Boolean OR keeps the variants in one search where the tool supports it.

> **Warning:** Result caps truncate a broad ICP without an error message. Sales Navigator shows up to 2,500 results per search, Apollo stops at 100 pages, and Clay's Find Companies returns at most 10,000 companies. A list that looks complete may be only the top of the sort order. Split the search until each part sits under the cap.

Step seven matters most for large markets. A cap does not reject the search. It quietly shows the first slice and hides the rest, so a list of exactly the cap figure is a sign to split, not a finished list.

Each tool section below follows the same order. It covers the starting point, the filters that carry the rules, the exclusion mechanism, caps and limits, plan gates where documented, the cost driver where documented, and where the tool fits.

## How To Write Exclusion Rules For Prospecting

Keep permanent disqualifiers apart from suppressions, because a suppression such as an opt-out is a legal duty, and the FTC's CAN-SPAM guide states that [the law makes no exception for business-to-business email](https://www.ftc.gov/business-guidance/resources/can-spam-act-compliance-guide-business) and that opted-out addresses cannot be sold or transferred.

An exclusion list mixes three kinds of rule, and each kind has a different owner and a different review cycle.

**Three kinds of exclusion rule**

| Kind | Examples | Changes when |
| --- | --- | --- |
| Permanent disqualifier | Competitors, wrong model, unserved regions | The ICP changes |
| Operational suppression | Customers, open deals, recently contacted | Every CRM sync |
| Legal suppression | Opt-outs, objections | Kept on file |

**Permanent disqualifiers** describe companies that will never fit. They belong in the saved search itself, as excluded industries, countries or company names.

**Operational suppressions** change every week. Current customers, accounts with an open opportunity, people contacted in the last cycle and previous downloads all belong here. Feed them from the CRM on every run instead of typing them into the search once.

**Legal suppressions** are duties, not preferences. The FTC's guide adds that opt-out requests must be honoured within 10 business days, and that an opted-out address cannot be sold or transferred, even as part of a mailing list, except to a company hired to help you comply.

In the UK, the ICO says [the PECR rule on direct marketing by email does not apply to corporate subscribers](https://ico.org.uk/for-organisations/direct-marketing-and-privacy-and-electronic-communications/business-to-business-marketing/), but sole traders and some partnerships count as individual subscribers and need specific consent or the soft opt-in. A filter on company type does not settle that question, so tag small firms and sole traders for a manual check.

The ICO also says a named business contact is personal data under the UK GDPR, even in a business capacity. It advises screening new B2B marketing lists against a do-not-email list and keeping objectors on a suppression list rather than deleting them. Deleting an objector removes the record that stops you contacting them again.

The tools in this guide carry exclusions in different ways:

- **UpLead** has an EU/EEA toggle, uploaded lists of emails, company phones or URLs, Salesforce connected lists, and previous downloads excluded by default.
- **Apollo** removes companies already on your lists with the list filter's Exclude option.
- **Ocean.io** excludes known companies by domain with its Company Lookup filter.
- **Clay** excludes up to three sets of companies or people, each needing a domain or profile URL.

Exclusion lists have ceilings. UpLead counts every criterion, exclusions included, toward a 200,000 cap, and Clay allows 300,000 companies across three exclusion sets. A large customer and suppression file may need splitting, or suppression after export, to fit.

This section is general information, not legal advice.

## Building ICP Filters In Apollo

Apollo carries an ICP as a saved company search plus a persona, uses the list filter's Exclude option to drop companies already on your lists, and stops results at 100 pages, so a broad ICP needs narrower filters before export.

**Starting point:** Save a company search alongside a list. Apollo then combines a persona, a set of job titles, functions, departments or locations, with the list filter so people searches run inside those accounts.

**ICP filters that carry the rules:** Company filters cover industry, number of employees, technologies, job postings and buying intent. A companies search limits results to businesses.

**Exclusion mechanism:** The list filter's Exclude option removes companies already on your lists. The location filter can exclude a country, such as the United States, for regions you cannot serve.

**Caps and limits:** Results stop at 100 pages, and Apollo's advice is to add filters to narrow them. The Net New and Total tabs cap how many contacts you can select at once, while the Saved tab allows bulk actions on up to 50,000.

**Plan gates and permissions:** Apollo states some plans include more filters than others, and advanced filters may require an upgrade. Saved searches can be shared with teammates.

**Where it fits:** Teams that want account filters, the persona and saved lists in the tool reps already use, with an ICP narrow enough to sit under the page cap. A very broad market needs splitting before it works here.

**List filter with Exclude** (documented): Apollo's help documentation describes a list filter that can include or exclude companies already saved to your lists, which keeps customers and worked accounts out of a new search.

For coverage and cost against other vendors, see the [B2B contact databases compared](https://growthtribune.com/best-b2b-contact-databases/) side by side.

## Building ICP Filters In LinkedIn Sales Navigator

In LinkedIn Sales Navigator, save the companies that pass your account filters as an account list and then run lead filters inside it, because a single search shows up to 2,500 results and revenue and technology filters rest on weaker data.

**Starting point:** An independent practitioner playbook recommends account filters for size, industry, countries and growth signals, adding adjacent industries because companies are often misclassified. Save qualifying accounts to an account list, then apply the Account lists lead filter.

**ICP filters that carry the rules:** Lead search includes Company headquarters, Current Company, Past Company, TeamLink connections of, Saved Leads, Saved Accounts and Previously Viewed. The CRM filter is named People in CRM. The search bar accepts AND, OR, NOT, quotation marks and parentheses, which keeps every title variant in one search.

**Exclusion mechanism:** LinkedIn's community guidance lists exclusion on Company, Seniority level, Title, Function and School for leads, and on Geography and Industry for leads and accounts, applied with a block icon. The independent playbook notes most filters are include-only, and saved lead lists can act as exclusions to find people not yet contacted.

**Caps and limits:** The results page shows up to 2,500 results per search. Up to 50 lead searches and 50 account searches can be saved, and saved searches can send alerts. The playbook treats self-reported annual revenue, technologies used and the lead profile industry filter as unreliable.

**Where it fits:** Teams whose buyers are active on LinkedIn and who prospect person by person. It suits account-list workflows well, and suits revenue or technology rules poorly.

**Account lists as a lead filter** (documented): Sales Navigator's lead search can be limited to a saved account list, so lead filters run only inside companies that already passed the account rules.

## Building ICP Filters In Ocean.io

Ocean.io starts from your best customers rather than a filter list, ranking lookalikes of up to 10 seed domains as A, B or C grades, and then narrows them with AI-assigned industry tags you can include or exclude.

**Starting point:** A lookalike search with seed domains. Ocean recommends starting with one to three strong seeds. Results are graded A, B or C, with C and above shown by default, and the grade works as a cutoff.

**ICP filters that carry the rules:** The Industries filter uses AI-assigned categories and tags read from company websites and LinkedIn pages. It can include or exclude, and can require Any of or All of the chosen tags. Ocean recommends adding Industries after the lookalike search to refine or exclude.

**Exclusion mechanism:** The Company Lookup filter includes or excludes companies by name or domain, pasted in or uploaded as a CSV with a Domain column.

**Caps and limits:** Result order is set by relevance and cannot be changed manually. Lookalike search struggles with websites that have little text or are mostly visual, generic website wording gives broader results, and results may need refinement on the first try.

**Plan gates and permissions:** Company Lookup is limited to paid plans. The uploaded list is not saved automatically, so save the search or the exclusion list is lost.

**Cost driver:** Adding companies to Company Lookup is free, loading a row costs 0.2 credits, and export after loading costs nothing extra. The Growth plan includes 60,000 credits a year, so loaded rows drive spend.

**Ocean.io pricing**

| Plan | Price | Billing | Notes |
| --- | --- | --- | --- |
| Growth | $250 | per month, billed annually |  |
| Revenue Navigator | Custom pricing | per custom |  |

**Where it fits:** Teams whose ICP is easier to show by example than to describe in fields. Teams selling into image-heavy or thin-website markets will see broader results.

**Company Lookup exclusion** (documented): Ocean.io's help centre describes Company Lookup as an include or exclude filter by name or domain, with the uploaded list kept only when the search is saved.

## Building ICP Filters In UpLead

UpLead turns ICP rules into NAICS, SIC, headcount, revenue and technology filters, excludes previous downloads by default, and caps all criteria, exclusions included, at 200,000 entries per search.

**Starting point:** Company filters for industry category, NAICS-based industry name, industry tags, SIC and NAICS codes, type of business, Fortune ranking, subsidiaries, employees, revenue, technologies, funding and year founded.

**ICP filters that carry the rules:** Contact filters cover job title, function and subfunction, management level, skills and location down to zip code and metro area.

**Exclusion mechanism:** Keyword exclusions (includes or exact) on names, titles, industries, cities, states and zip codes. Industry exclusions, an EU/EEA toggle that removes all EU/EEA contacts and companies, uploaded lists of emails, company phones or company URLs, and Salesforce connected lists. Previous downloads are excluded automatically by default.

**Caps and limits:** Total criteria across all filters, exclusions included, are capped at 200,000, and list entries count toward the cap. Search Lists are single-column CSV uploads of one type (state, zip, NAICS code, job title, company URL or exclusions).

**Where it fits:** Teams that want firm industry codes and a built-in EU/EEA exclusion. A very large suppression file has to fit under the shared criteria cap.

**Exclusions** (documented): UpLead's help centre documents exclusions by keyword, industry, uploaded list, Salesforce list and an EU/EEA toggle, with previous downloads excluded by default.

For a side-by-side on coverage and export terms, see [how ZoomInfo and UpLead compare](https://growthtribune.com/zoominfo-vs-uplead-which-fits-your-team/).

## Building ICP Filters In Clay

Clay lets one search combine firmographic, technographic and AI-classified filters with up to three exclusion sets of 300,000 companies in total, but returns at most 10,000 companies per search, so large ICPs are split by segment.

**Starting point:** Find Companies, with industries (include or exclude), size, revenue brackets, funding, company type, location, keywords with include, exclude and exact phrase, AI filters (industries, subindustries, revenue streams, business type), technographics and a free-text semantic description that ranks fit.

**ICP filters that carry the rules:** Find People matches titles with "is similar to", "contains" or "is exactly". "Contains" excludes synonyms and "is exactly" is case-sensitive. It filters by company size, industry, location, experience and a company list from a table.

**Exclusion mechanism:** Up to three sets of companies from Clay tables, CSVs or manual lists, up to 300,000 in total and 100,000 per source. Each excluded company needs a domain or a professional profile URL. Find People excludes up to three sets (300,000 people) that need profile URLs.

**Caps and limits:** A company search returns at most 10,000 results. Find People returns up to 50,000 records per search and 100 people per company. Special characters and stopwords are removed even inside quoted phrases.

**Cost driver:** Clay plans are metered in actions and data credits per month, and paid plan cards list starting allowances that can be expanded, so enrichment volume drives cost. The Ocean.io integration inside Clay finds company lookalikes from a target domain, and each lookalike result consumes one credit.

**Clay pricing**

| Plan | Price | Billing | Notes |
| --- | --- | --- | --- |
| Launch | $185 | per month |  |
| Growth | $495 | per month |  |
| Enterprise | Custom pricing | per custom |  |

**Where it fits:** Teams that want every filter and exclusion in one table they can enrich and reuse. Teams that rely on exact title matching should test their title sets against the "contains" and "is exactly" rules.

**Exclusion sets** (documented): Clay University documents up to three exclusion sets per search, each company identified by domain or profile URL, with limits per source and in total.

## How To Check A Filtered Prospect List Before Outreach

Before a filtered list reaches reps, score a random sample of its accounts against each ICP rule by hand, confirm every suppression list was applied, and check that contacts span the buying group, which Gartner puts at five to 16 people across as many as four functions.

The checks below are the success measures for the filter set. Run them on every new search and on every rule change.

- [ ] **Score a random sample by hand.** Pull a random set of accounts (20 to 30 is a workable rule of thumb) and mark each one pass or fail against every rule. Note which rule let each bad fit through.
- [ ] **Tighten the rule that failed.** Re-run the search with that one rule changed, then sample again.
- [ ] **Check contacts per account.** Each account should hold contacts across several functions in the buying group, not one title repeated.
- [ ] **Confirm every suppression ran.** Look for current customers, open opportunities, opted-out contacts and previous downloads in the output.
- [ ] **Check no search hit its cap.** A count equal to the tool's limit (2,500 in Sales Navigator, 100 pages in Apollo) means the list was truncated.
- [ ] **Track outcomes by segment after sending.** Compare reply, meeting and win rates by ICP segment and change one filter at a time.

Reading 20 or 30 company websites against the rule sheet shows which rule let a poor fit through, such as a misclassified industry or an out-of-date headcount. The goal is not a fixed pass rate. It is knowing which rule fails and why.

Contacts per account matter as much as account fit. A list of fit accounts with one finance title each reaches one member of a group of several. Check the spread of functions before outreach, not after a sequence stalls.

The same frozen sample can serve twice. The method used to [test coverage against a frozen ICP sample](https://growthtribune.com/how-to-evaluate-b2b-database-coverage-before-buying/) before buying a database also works after the filters are built, as a fixed benchmark for each new version of the rule sheet.

> **Tip:** Keep a frozen sample of 20 to 30 accounts you know are good fits and run every new filter version against it. If the new version drops a known fit, the rule that removed it needs rewriting.

Outcome data closes the loop. Reply and meeting rates show which segments respond. Win rates by segment show which rules describe buyers rather than responders. Change one filter at a time, so each result can be traced to its cause.

## How To Fit ICP Filters Into A B2B Prospecting Workflow

Treat the ICP filter set as a saved, named search with an owner, so that Apollo can re-run and share it, Sales Navigator can alert on new matches, and suppression lists from the CRM are re-applied on every run. Without an owner, the rules drift and nobody notices.

Filters are not a one-off build. They live inside a workflow with inputs, outputs and an owner at each stage.

**Where ICP filters sit in a prospecting workflow**

| Stage | Input | Output | Owner |
| --- | --- | --- | --- |
| Rule sheet | Closed-won and lost accounts | Testable ICP rules | RevOps |
| Account search | Rule sheet | Saved account list | RevOps |
| Suppression feed | CRM customers, opt-outs | Exclusion lists | RevOps |
| Person search | Saved account list | Contacts by role | SDR manager |
| Sample check | Random accounts | Pass or fail notes | SDR manager |
| Outreach | Checked list | Replies and meetings | SDRs |
| Review | Results by segment | Rule changes | RevOps and sales |

Each tool keeps the reusable part in a different place:

- **Apollo** saves company searches alongside a list and shares saved searches with teammates.
- **Sales Navigator** saves up to 50 lead and 50 account searches, and saved searches can send alerts on new matches.
- **Ocean.io** keeps an uploaded Company Lookup list only when the search is saved, so save before you leave the page.
- **Clay** uses a table as the reusable company list, and exclusion sets can come from other tables.

Some teams chain these tools, building the account list in one and enriching it in another. The [no-code prospect list building tools](https://growthtribune.com/best-no-code-prospect-list-building-tools/) cover how those chains are set up. Whatever the stack, the rule sheet stays the source of truth and each tool's search is a copy of it.

Two habits keep the workflow honest. First, re-apply suppression lists from the CRM on every run, because customers and opt-outs change between runs. Second, version the rule sheet and name each saved search with that version, so a change in results can be traced to a change in rules.

A refresh cadence follows from your outcome data. When segment results shift, review the rules. When the market or the product changes, rewrite them from new closed-won accounts.

## Frequently Asked Questions

An ICP filters companies and a buyer persona filters people inside them, so in tools such as Apollo the persona of titles, functions and locations is applied on top of a saved company list.

**How many criteria should an ICP filter set have?**

Keep the ICP filter set to a short list of testable rules, with three to seven as a working rule, each tied to one field. Fewer rules let poor fits through, while many more shrink the list and hide which rule removed a good account. Put untestable traits in manual checks instead.

**What is the difference between an ICP and a buyer persona when filtering?**

An ICP filters companies and a buyer persona filters people inside them. In Apollo, for example, a persona of titles, functions and locations runs on top of a saved company list. Write the account rules first, then a persona for each role in the buying group, never one title alone.

**Do I need consent to email business contacts from a filtered list?**

It depends on where the recipient is. The FTC says CAN-SPAM makes no exception for business-to-business email, so opt-outs always apply. In the UK, the ICO says the PECR email rule does not cover corporate subscribers, but sole traders and some partnerships need consent or the soft opt-in. This is general information, not legal advice.

**How often should prospecting filters be reviewed?**

Review prospecting filters whenever reply, meeting and win rates by ICP segment give you enough results to compare, and change one filter at a time so each effect stays readable. Re-apply suppression lists on every run, and give the shared rule sheet and saved searches one owner who signs off each change.

## Sources

- [Gartner Sales Survey Finds 74% of B2B Buyer Teams Demonstrate Unhealthy Conflict During The Decision Process](https://www.gartner.com/en/newsroom/press-releases/2025-05-07-gartner-sales-survey-finds-74-percent-of-b2b-buyer-teams-demonstrate-unhealthy-conflict-during-the-decision-process), gartner.com
- [CAN-SPAM Act: A Compliance Guide for Business](https://www.ftc.gov/business-guidance/resources/can-spam-act-compliance-guide-business), ftc.gov
- [Business-to-business marketing | ICO](https://ico.org.uk/for-organisations/direct-marketing-and-privacy-and-electronic-communications/business-to-business-marketing/), ico.org.uk

---
Source: https://growthtribune.com/how-to-turn-an-icp-into-prospecting-filters/