How-To Guides16 min read·Updated 2026-04-30

How to Use Clay for B2B Prospecting

Clay is not a database. It is a workflow platform that pulls from 100+ providers and runs AI research. Here is how to use it to build prospect lists that actually convert.

RB

Rees Bayba

Founder, Astra GTM

TL;DR

  • Clay is a workflow platform, not a database. It pulls from 100+ providers (Apollo, BuiltWith, Clearbit, LinkedIn, RocketReach) and runs AI research on each row.
  • The workflow: build company list → enrich company data → find contacts → enrich emails → AI research → score and filter → push to sequencer.
  • Claygent (Clay's AI agent) browses company websites and LinkedIn to extract custom insights that no database provides -- recent hires, tech triggers, news, product launches.
  • Waterfalls are how you avoid wasting credits: try Provider A first, only run Provider B if A found nothing. Set fallback logic before running any enrichment at scale.
  • Cost structure: most data enrichments cost 2-5 Clay credits. AI enrichments (Claygent) cost 5-10 credits. One credit is approximately $0.01.

Clay is not a lead database. Companies that use it as a more expensive Apollo are wasting 80% of what it can do. Clay is a workflow platform: you define a series of enrichment steps, and Clay pulls data from whichever of its 100+ integrated providers has the best answer for each row. Then it runs AI research on top of that data. The output is a prospect list that knows things about each company that no static database could tell you.

Step 1: Build Your Target Company List

Every Clay workflow starts with a list of companies. Three ways to get that list into Clay, depending on what you have.

Import from CSV

If you already have a list of target companies or domains from any source -- a manual research list, a trade show attendee list, an export from your CRM -- import it as a CSV. Clay will create one row per company and you enrich from there.

Connect Apollo or LinkedIn Sales Navigator

Clay's Apollo integration and LinkedIn Sales Nav integration let you run a search directly inside Clay and pull the results as rows. Apollo is better for broad ICP searches (industry + headcount + geography). Sales Navigator is better for searches that require LinkedIn-native filters (function, seniority, company growth rate, recent activity).

Use Clay's native company search

Clay has a built-in company search powered by its own data layer. Filter by industry, headcount, revenue, tech stack, geography, and keywords. Less flexible than Apollo or Sales Nav but useful when you do not have either subscription.

Step 2: Enrich Company Data

Once you have a company list, enrich it with the signals you will use for scoring and personalization. Four enrichments cover most use cases.

BuiltWith -- tech stack

BuiltWith tells you what technologies a company's website runs on. Marketing automation tools, CRMs, analytics platforms, payment processors, CDNs. If your product competes with or integrates with specific tools, BuiltWith lets you filter to companies already using them or identify companies stuck on legacy tools. Add BuiltWith as a column, pass in the company domain, and get back a structured list of detected technologies.

Clearbit -- firmographics

Clearbit provides structured company data: industry, sub-industry, headcount, estimated revenue, funding stage, description, social profiles. More reliable than scraping LinkedIn for company size. Use Clearbit data to populate the firmographic columns you will filter and score on.

LinkedIn -- headcount and growth rate

Clay's LinkedIn enrichment pulls current headcount and 6-month headcount growth rate. These two signals together reveal whether a company is actively scaling (positive hiring signal) or contracting (negative signal). Companies growing headcount by 20%+ in 6 months are often in buying mode for tools that support that growth.

Clay's own enrichment -- website and description

Clay's native enrichment scrapes the company website and extracts a clean one-paragraph description. Useful for feeding into Claygent prompts (Step 5) and for filtering out companies that don't match your ICP based on what they actually do, not just their SIC code.

Step 3: Find Contacts

After enriching companies, find the specific people you want to reach. Clay's people search lets you filter by job title, seniority, function, and keywords within a company. For most B2B outbound, you are looking for one or two buyer personas per company.

Use a waterfall across multiple providers

No single data provider has complete coverage. Build a waterfall: try Apollo first (best coverage for US companies), then RocketReach if Apollo returns nothing, then Lusha as a final fallback. Clay's waterfall column type handles this automatically -- it runs each provider in sequence and stops when it finds a result, so you only spend credits when necessary.

  • Provider 1: Apollo -- strongest US coverage, 275M+ contacts
  • Provider 2: RocketReach -- strong for senior executives and international markets
  • Provider 3: Lusha -- strong for EMEA, sometimes finds contacts the other two miss
  • Stop condition: as soon as any provider returns a LinkedIn URL or contact record, stop the waterfall

Step 4: Enrich Contact Emails

Finding a contact name and LinkedIn profile is not enough. You need a verified email address. Run another waterfall specifically for emails, with different providers optimized for email finding.

ProviderStrengthClay creditsBest for
HunterDomain-level pattern matching2 creditsProfessional emails on small-medium domains
DropcontactGDPR-compliant, strong European coverage4 creditsEuropean contacts
ApolloHigh volume, broad coverage2 creditsUS/NA B2B contacts
RocketReachStrong executive coverage5 creditsVPs, C-suite, senior ICs

After the email waterfall, verify every result with a separate verification step. Clay integrates with NeverBounce and ZeroBounce for verification. Do not export unverified emails to your sequencer -- invalid emails burn sender reputation fast.

Step 5: Run Claygent AI Research

Claygent is where Clay separates from every other tool. It is an AI agent that browses company websites, LinkedIn pages, and recent news for each row in your table and extracts custom insights based on whatever prompt you write. No static database has this data. It has to be researched in real time.

What to ask Claygent

  • Recent hires: 'Has this company hired a VP of Sales or CRO in the last 90 days? If yes, return their name and LinkedIn URL.'
  • Tech triggers: 'Does this company's job postings mention Salesforce, HubSpot, or Outreach? List which tools appear most.'
  • Growth signals: 'Has this company announced a new product, a new market, or a funding round in the last 6 months? Summarize in one sentence.'
  • Pain signals: 'Based on their website and recent LinkedIn posts, what operational challenge does this company appear to be working through?'
  • Personalization hooks: 'What is one specific thing about this company that would be a relevant opener in a cold email about [your product category]?'

Claygent costs 5-10 credits per row depending on how complex the research task is. Run it after filtering -- you do not want to spend Claygent credits on companies that will not pass your scoring threshold. Score first (Step 6), then run Claygent only on companies that score above your cutoff.

100+
data providers available inside Clay

Apollo, LinkedIn, BuiltWith, Clearbit, Bombora, Crunchbase, ZoomInfo, Hunter, RocketReach, Lusha, and 90+ more. The value is not any single provider -- it is the ability to waterfall across all of them and use AI to synthesize the outputs.

Step 6: Score and Filter

Before you export or push to a sequencer, score each company and filter to your best-fit accounts. Clay's formula columns let you build a simple point-based score.

Simple ICP score formula (Clay formula column)

Score = 0 + 1 if funded (funding_stage is not null) + 1 if tech match (builtwith_technologies includes your target tool) + 1 if hiring signal (headcount_growth_6mo > 10%) + 1 if company size match (headcount between 50 and 500) + 2 if Claygent found a recent trigger (ai_trigger is not null) Filter to score >= 3 before export.

A tight score filter turns a raw list of 5,000 companies into 400 high-confidence targets. Your sequencer performance is almost entirely determined by the quality of the list going in. The scoring step is where you earn your reply rate.

Step 7: Export or Push to Your Sequencer

Clay connects directly to Instantly, SmartLead, HubSpot, Salesforce, and most other sales tools. You can push rows that pass your score threshold directly into a campaign, or export as CSV and import manually.

  • Instantly and SmartLead: Clay has native integrations that push contacts directly to campaigns
  • HubSpot: native integration pushes to contacts and companies with field mapping
  • Outreach and Salesloft: push via the CSV export or Zapier for platforms without native connectors
  • Always verify emails before pushing to any platform -- invalid emails and spam traps damage sender reputation

Understanding Clay Credits

Clay's pricing is credit-based. Each enrichment action costs a certain number of credits. Understanding the credit cost of each step prevents expensive surprises.

ActionCredits usedApprox. cost
Apollo company enrichment2 credits~$0.02
BuiltWith tech stack lookup2 credits~$0.02
Clearbit firmographics3 credits~$0.03
Hunter email find2 credits~$0.02
RocketReach email find5 credits~$0.05
Claygent research (standard)5-10 credits~$0.05-0.10
NeverBounce email verification1 credit~$0.01
Clay native company search0 credits (free)$0

How to avoid credit waste

  • Build waterfalls with stop conditions -- never run all providers on every row
  • Filter first, enrich second -- apply basic ICP filters before running expensive enrichments
  • Run Claygent last, after scoring -- only research companies that pass your score threshold
  • Test on 10-20 rows before running a full batch of 1,000+
  • Use Clay's 'run only on empty' setting to avoid re-enriching rows that already have data

Frequently asked questions

Is Clay worth it compared to just using Apollo?

Apollo is a database. Clay is a workflow platform that can use Apollo as one of its sources. If you only need a list of contacts with emails, Apollo is cheaper and simpler. If you want to enrich companies with tech stack data, run AI research on each account, build signal-based scoring, and push enriched data directly to your sequencer, Clay does things Apollo cannot. The value grows with the complexity of your ICP and personalization requirements.

How much does it cost to build a list of 1,000 companies in Clay?

For a standard workflow (company enrichment + contact finding + email waterfall + basic Claygent research), expect 15-25 credits per row. At 1,000 rows, that is 15,000-25,000 credits. Clay's Starter plan includes 2,000 credits/month ($149). The Explorer plan includes 10,000 credits/month ($349). Most teams running regular prospecting workflows land on the Explorer or Pro plan.

What is the best way to find contact emails in Clay?

Use a waterfall: Hunter first (fastest, cheapest), then Apollo, then RocketReach or Lusha if the first two find nothing. Set the waterfall to stop as soon as any provider returns an email. After the waterfall, run NeverBounce or ZeroBounce verification on every email. Do not push unverified emails to any campaign platform.

Can Claygent access LinkedIn data?

Claygent browses public LinkedIn company pages (not Sales Navigator) and can extract public information -- company description, recent posts, employee count, recent job postings. It cannot access private profiles or Sales Navigator-gated data. For LinkedIn profile-level data (contact work history, skills, activity), use Clay's dedicated LinkedIn enrichment column, which pulls from LinkedIn's data via Clay's own infrastructure.

How do I avoid running out of Clay credits mid-workflow?

Three practices: (1) Always test on 20 rows before running the full batch. (2) Build waterfalls with strict stop conditions so expensive fallback providers only run when cheaper ones fail. (3) Use Clay's credit cost preview feature before running any enrichment on a full table -- it shows estimated credit consumption before you commit. Check your credit balance before any large run.

Does Clay integrate with my CRM to avoid contacting existing customers?

Yes. Clay connects to HubSpot, Salesforce, Pipedrive, and most major CRMs. You can pull your existing contacts or accounts into Clay and use a filter to exclude any domain that already exists in your CRM before exporting. Build this deduplication step as the last filter before export or push -- it prevents the embarrassing situation of cold emailing your own customers.

Want this built for your team?

We implement these systems end-to-end. First sends within 14 days.