Lead Sourcing: How to Find Net-New B2B Leads Not Already in Your CRM [2026]

Lead sourcing is the step where you decide which people enter your pipeline, and net-new sourcing means finding buyers who are not already sitting in your CRM. The order that works is dedupe first, source second, enrich last: export what you already hold, subtract it from the target set, then spend money only on what survives. Cleanlist covers the source-and-enrich half of that loop through People Search and Company Search (0 credits to run), lead lists (0.5 credits per lead saved) and bulk waterfall enrichment (1 credit for a verified email, 10 for a direct dial, 11 for both). This page rates eight sourcing channels on cost, freshness and legal risk.

Last updated: August 15, 2026. Cleanlist prices, credit costs and product capabilities on this page reflect the product as of that date. Third-party prices and legal citations were checked on August 15, 2026 and are linked inline.

What actually counts as a net-new lead?

A net-new lead is a person who does not already exist as a contact record in your CRM, has no open or closed-lost opportunity attached, and is not on a suppression list. Most teams skip three of those four checks, which is why "net-new" campaigns routinely re-mail existing customers.

Watch the term itself. Inside Salesforce reporting, "net new" often means a campaign-attribution bucket rather than a sourcing job, so the request can mean two different things depending on who is asking. This page treats it as the sourcing job: find people you have never held. Cleanlist sits on the sourcing side of that line. It finds people, resolves how to reach them, and pushes them into the CRM or sequencer that owns the record afterwards.

Why do you deduplicate against your CRM before you source anything?

Because every duplicate you enrich is money spent on a contact you already own. At Cleanlist rates a row costs 0.5 credits to save plus 1 to 11 credits to enrich, so the 400 duplicates hiding in a 2,000-row list that is 20% overlapped burn 600 credits at the email-only rate and 4,600 at the full-contact rate, before a single email goes out. Deduplicating afterwards does not refund any of it.

The mechanical version: export contacts and leads from your CRM with email and company domain, keep them in a single suppression file, and subtract on email first and on company_domain plus job_title second. Inside Cleanlist there are two supports for this. The CRM Contact Lookup Smart Agent column checks a connected CRM and flags rows you already hold, and smart agent columns run 0.5 to 3 credits depending on the agent. The Playbook Builder on the $599 Scale plan carries a dedupe step inside a repeatable import, dedupe, enrich, verify and export flow. Neither one suppresses your CRM automatically at search time. The subtraction stays a step you run before you spend.

Where do net-new B2B leads actually come from in 2026?

Eight channels, and a healthy sourcing motion runs three or four of them rather than betting on one. The verdicts below are about the channel, not about any vendor.

The pattern worth noticing: the cheapest channel per record is usually the stalest, and the freshest channel (your own inbound) is the one you cannot scale on demand.

What does structured people and company search actually get you?

Structured search is the channel where you write your ICP as filters and get back matching profiles. Cleanlist People Search exposes about 24 distinct filter fields across seven groups: title with include and exclude, the seniority ladder from Owner-Partner down to Entry, years in role, headcount in eight bands from 1-10 to 10,001+, contact and company-HQ location with exclusions, industry on the LinkedIn taxonomy, skills, experience, recently changed jobs, plus education and past titles and companies.

Running a search costs 0 credits in Cleanlist, which changes how you work: you iterate the filter set and watch the result count move before spending anything. What comes back for free is the profile side (name, title, company, industry, headcount band, location, LinkedIn URL). Emails and phones are resolved at enrichment time through a multi-provider waterfall, not held in advance.

Is exporting from LinkedIn Sales Navigator a viable sourcing channel?

Sales Navigator is the strongest targeting surface in B2B and the weakest output. LinkedIn's own plan comparison, checked August 15 2026, lists Core at US$119.99 per month (US$1,079.88 per year) and Advanced at US$159.99 per month (US$1,799.88 per year), and lists neither verified work email addresses nor direct-dial phone numbers among plan features (LinkedIn Sales Solutions).

So the channel produces a targeted set of names with no way to contact them. The workable pattern is to build the segment in Sales Navigator, export the list, then enrich it: Cleanlist takes a Sales Navigator export or a column of LinkedIn URLs and resolves work emails at 1 credit and direct dials at 10. Method matters more than tooling here, because LinkedIn's User Agreement prohibits scraping the results, and data pulled from a logged-in session sits worse in a legitimate-interests assessment than data sourced through providers with published compliance documentation. The compliant export methods are laid out in the Sales Navigator export guide.

When is the Chrome extension the right sourcing tool?

When the segment is small, hand-picked and impossible to express as filters. Sourcing "the twelve people who spoke at this conference" or "everyone who commented on this post" is a browsing job, and no filter set will reproduce it. The Cleanlist Chrome extension captures profiles from LinkedIn and Sales Navigator pages you are already looking at and drops them into a lead list, where the same enrichment prices apply.

The honest limit is throughput. An extension moves at the speed of a human scrolling, so it is a poor fit above a few hundred rows. Run it for named-account committees where precision is the point, and reserve People Search for the top-of-funnel sweep.

How do you find companies similar to your best customers?

Start from your closed-won accounts, extract the firmographic pattern they share, then search on that pattern. Cleanlist supports the shortcut two ways: a Find Similar Companies Smart Agent column in the app, and POST /search/companies/similar on the REST API v2. Here is the honest description, since the category tends to oversell this feature. The Cleanlist similar-companies result is synthesized from firmographic similarity. No proprietary lookalike model is trained on your revenue data, and no ranked prediction is being made. Treat the output as a candidate set to review.

The workflow that beats the button: export your last 20 closed-won accounts, write down what they share on the dimensions Cleanlist Company Search actually indexes (industry, headcount band, HQ location, company type, year founded), then run that as a search. Company Search costs 0 credits, a company enrichment costs 1. Then run People Search filtered by company name plus seniority to pull the buying committee inside each account.

How do you find companies that just raised funding?

Not through a Cleanlist filter. Say the constraint first, because a campaign built on a filter that does not work is a wasted week. Cleanlist People Search has no funding-stage filter at all. On the REST API, POST /search/companies accepts a funding_stage key in its schema, and that key returns zero rows today, so treat it as unusable rather than as a working filter. There is also no revenue filter and no technographic filter anywhere in Cleanlist, because none of those are indexed fields.

The real approach is two-step: get the funding list from a source that actually tracks funding, then enrich it. For US private placements the primary source is free and legally required, because 17 CFR 230.503(a)(1) requires an issuer relying on Rule 504 or 506 to file a Form D notice "no later than 15 calendar days after the first sale of securities in the offering" (Cornell LII). Form D filings are public on the SEC's EDGAR full-text search, they carry the issuer name and address, and the 15-day rule means the feed runs close to the event. Beyond that, funding databases, VC portfolio pages, press release wires and hiring surges all work as inputs.

Once you have the company list, Cleanlist takes over: paste the domains, run People Search by company name plus seniority to pull the roles you sell to, then enrich. The funding signal comes from outside, and the contact data comes from the waterfall.

Are public lists, directories and communities worth the time?

Yes for niches no filter set can express, and no for anything you could have searched for. If your ICP is "Shopify apps with a public changelog" or "agencies in a specific accreditation register", a directory beats any people-search product, because membership in that register is the ICP and no B2B index carries it as a field.

The cost is time rather than money, and the output is a company list rather than a contact list. The two-step handles that: harvest the names or domains, paste them into Cleanlist Company Search or People Search filtered by company name, and let the waterfall resolve the people. The failure mode is treating a scraped directory page as a finished lead list. Directory contact fields are usually a generic info@ inbox, which is the worst possible target for a cold sequence.

What about inbound signals, referrals and partner lists?

These are the highest-converting sources and the ones you cannot summon on demand. Inbound (form fills, free tool users, content downloads) converts at multiples of cold outbound because the person raised a hand, and referrals carry borrowed trust. Both belong at the top of the working queue before any sourced row gets touched.

The job on these is enrichment rather than discovery, and it is where Cleanlist does its cleanest work: a form fill usually arrives as an email address and a company name with no title, seniority or headcount, so routing it is guesswork. Enriching an inbound record at 1 credit fills the firmographics that scoring needs. One boundary worth stating: consent does not travel automatically with a partner list. The lawful basis you can rely on is the one their collection notice supports.

Should you buy a prepackaged B2B lead list?

Buying is the fastest way to a large file and the fastest way to a damaged sending domain. Four questions decide it. Ask them in writing, and treat an unanswered one as a reason to walk.

1. When was each record last verified, per record, with a date in the file? "Recently" is not an answer.
2. Where did the data originate, and can the seller name the collection basis? You inherit that basis when you send.
3. Is the list exclusive, or has it been sold to everyone in your category? Non-exclusive lists arrive pre-fatigued.
4. Is there a bounce guarantee with an actual remedy, or only a stated accuracy figure?

The structural problem is that a bought list is a snapshot and B2B contact records decay continuously as people change jobs, so a list sold in January and sent in June has drifted. Cleanlist resolves each address at the moment you enrich rather than at the moment somebody collected it, which is why it prices per resolved record instead of per file. If you do buy, verify the whole file before the first send.

What does it cost per usable lead to build a list instead of buying one?

Work it from the credit prices, which are published and fixed. In Cleanlist, search costs 0 credits, saving a newly added lead to a list costs 0.5, a verified email costs 1, a direct dial costs 10, both together cost 11, a company enrichment costs 1, and a CRM or sequencer sync costs 0.2 per lead.

Divide the plan price by the credit allowance to get a per-credit rate, then multiply by the credits a row consumes.

Annual billing takes 25% off those plan prices, and the free plan is $0 with 30 credits a month and no card. So a saved, email-enriched, net-new row lands between 6 and 8 cents at Cleanlist list prices, and a row with a direct dial attached lands between 46 and 61 cents. Compare that against whatever a broker quotes per record, then weight the broker's number by the share of the file you expect to discard on verification. Cleanlist's published specs are 98% for verified email accuracy and 85% for direct dial coverage, stated as company product specs rather than third-party test results.

How many leads do you actually need to book ten meetings?

Work backwards, and treat every rate below as an illustrative planning assumption rather than a Cleanlist measurement. Replace each one with your own historical numbers, because the answer swings by an order of magnitude between segments.

At Cleanlist prices that plan costs 2,455 saves at 0.5 credits (1,228) plus 2,455 email enrichments at 1 credit, roughly 3,683 credits, which fits inside one month of the $229 Pro plan. The number that moves the model hardest is meetings per delivered email, so measure yours before you buy volume. Sourcing 2,888 profiles against a 0.2% reply reality is how teams burn a sending domain and a quarter at the same time.

What do you do with leads that come back without an email?

Nothing is silently billed for a blank. Cleanlist enrichment is pay-for-results: the credits shown when you dispatch a job are a maximum reservation, and the actual charge can be lower or zero when no new data comes back. A row the waterfall cannot resolve stays in your list carrying the profile fields the search already returned.

Four things to do with those rows, in order of cost. Try a direct dial at 10 credits, which resolves a different data type through a different set of providers. Route them to LinkedIn touches, since you still hold the profile URL. Run an AI column at 0.5 to 3 credits to research something specific. Or drop them. What you should never do is guess a pattern like first.last@domain.com and send to it, because an unverified guess is the input that pushes a spam rate up and a sending domain down.

What is the compliance floor for EU and California contacts?

Cold B2B email is lawful in the US with conditions, and the conditions are jurisdictional rather than moral. In the US, 15 U.S.C. 7704 requires "a functioning return electronic mail address or other Internet-based mechanism, clearly and conspicuously displayed" that stays live "no less than 30 days after the transmission of the original message", opt-outs honoured within "10 business days", and "a valid physical postal address of the sender" (Cornell LII).

For EU and UK contacts, GDPR Recital 47 states that "the processing of personal data for direct marketing purposes may be regarded as carried out for a legitimate interest" (gdpr-info.eu). That is a documented balancing test you have to be able to produce, plus notice and an easy objection route. For California, the business-to-business carve-out is gone: Civil Code 1798.145(n)(3) states "This subdivision shall become inoperative on January 1, 2023" (California Legislative Information), so a California employee's work contact details are consumer personal information today.

Practically this is a source column and a date stamp on every row, because when someone objects you need to say where the record came from and when. Cleanlist stamps enrichment results per record, so that column gets filled at creation rather than reconstructed from memory.

What does a sourcing loop that compounds look like?

Weekly rather than quarterly, and small enough to actually run. Refresh the suppression file from the CRM. Add one sourced batch sized to what the sequencer can absorb. Enrich email-only first at 1 credit, and reserve the 11-credit full-contact spend for segments that already replied. Read the replies as ICP feedback, then tighten the filters for the next batch.

Weekly beats quarterly because of decay. A quarterly rebuild throws away the suppression history that stops you emailing the same person four times, and it re-sources contacts whose titles moved in the interim. The Cleanlist REST API v2 (Pro plan and above, $229/mo) and the MCP server both drive search, save and bulk-enrich, so the loop can be scheduled instead of remembered. Note the cap when you plan it: each API key is limited to 60 People Searches per UTC day.

FAQ

What is lead sourcing?

Lead sourcing is the process of identifying and collecting prospects who match your ideal customer profile, before any enrichment, scoring or outreach happens. It answers "which people enter the pipeline", and it sits upstream of list building, which answers "what does each row contain". In practice sourcing means picking channels (structured search, LinkedIn exports, directories, inbound, referrals, purchase) and running the dedupe pass that removes anyone you already hold. Cleanlist covers the structured-search and enrichment half of that: search costs 0 credits, saving a lead costs 0.5, and a verified email costs 1 credit.

How do I find leads that are not already in my CRM?

Export your CRM contacts and leads with email and company domain into a single suppression file, run your sourcing search, then subtract on email first and on company domain plus job title second. Do the subtraction before enrichment, because enriching a duplicate costs the same as enriching a net-new row. In Cleanlist the CRM Contact Lookup Smart Agent column flags rows already present in a connected CRM at 0.5 to 3 credits per row, and the Playbook Builder on the $599 Scale plan carries a dedupe step inside a repeatable flow. There is no automatic CRM suppression at search time, so treat it as a step you run.

Can Cleanlist filter for companies that just raised funding?

No. Cleanlist People Search has no funding-stage filter, and the funding_stage key accepted by the company search schema on the REST API returns zero rows today, so it is not a usable filter. There is no revenue filter and no technographic filter either, because those fields are not indexed. The working approach is to build the funded-company list from a source that tracks funding (SEC Form D filings on EDGAR are free and must be filed within 15 calendar days of the first sale under 17 CFR 230.503, plus funding databases and VC portfolio pages), then paste those domains into Cleanlist and enrich the buying committee.

Is it cheaper to buy a B2B lead list or build one?

Buying is cheaper per record at the moment of purchase and usually more expensive per usable record, because you pay for a snapshot and then discard the share that fails verification. Building through Cleanlist prices each resolved record instead: a saved, email-enriched row costs 1.5 credits, which is about 7 cents on the $229 Pro plan (5,000 credits) and 6 cents on the $599 Scale plan (15,000 credits). Full contact records with a direct dial run 11.5 credits, roughly 46 to 61 cents. The comparison that matters is cost per record that is still deliverable on send day, not cost per row in the file.

How many net-new leads should I source per month?

Size it to what your sequencer and your reps can actually work, then let the meeting math set the floor. Working backwards from ten meetings with illustrative assumptions (0.6% meetings per delivered email, a 3% bounce allowance, 70% of sourced rows resolving to a sendable address, 15% removed as duplicates) puts the requirement near 2,900 raw profiles a month. Measure your own reply and meeting rates before trusting that figure, because it moves by an order of magnitude across segments. Sourcing more volume than your sending infrastructure can absorb damages the domain that everything else depends on.