The Ultimate Clay API Guide for Cold Outreach Success [2026]

Most Clay guides will tell you the same thing: build a table, add enrichment columns, run the waterfall, push to Smartlead. That’s fine as far as it goes — but it misses the step that actually determines whether your campaign lands or kills your domain reputation before you’ve sent 200 emails.

The step everyone skips: your input data quality. Clay is an enrichment and personalization engine — it can’t fix a list that’s 20–30% stale before enrichment even starts. The standard advice is “verify emails in Clay.” But verification only tells you if an address bounces; it doesn’t tell you if the person still works there, or if the company even exists. If you’re pulling from a static database that was last refreshed 8 months ago, Clay is polishing data that was already degrading when you got it.

At Generect, we build and run Clay workflows daily — and the single biggest lever we’ve found isn’t the enrichment waterfall or the Claygent prompt. It’s starting with contacts fetched live, not bought from a static archive. This guide covers both: the Clay mechanics, and how to feed them with data that’s actually current.

Specifically, you’ll learn:

• What “Clay API” actually means (and the clever workaround for the fact that Clay doesn’t have one)
• How to structure tables so data flows without wasted credits
• The Generect → Clay workflow, step by step, with actual column names and API calls
• How Claygent works and how to run it without burning your Actions budget
• Which sending tools connect cleanly and how to wire them up
• The 9 pitfalls that kill Clay campaigns (including the data quality one nobody talks about)

What changed in Clay in 2026 (read this first)

If you read an older Clay guide, some of what you learned is now outdated. Here’s what actually changed since early 2026:

• New pricing structure (March 2026): Clay replaced Starter, Explorer, and Pro with Free, Launch ($185/mo), and Growth ($495/mo). The key change: they split usage into Data Credits and Actions — two separate buckets. Data marketplace costs dropped 50–90% across most providers in this overhaul, which significantly changes the cost math for enrichment-heavy workflows.
• Claygent is now a primary workflow tool: Claygent (Clay’s AI web-browsing agent) can visit a company’s pricing page, read a LinkedIn profile, or analyze any URL and return a structured answer — at scale, across your entire table. In 2026, teams use it as a first-pass research layer before any manual review.
• Intent signals are now central: Clay-powered campaigns that layer 2+ intent signals (job postings, funding events, leadership changes, tech stack shifts) see 2–4x higher reply rates than list-blasting with zero signals.
• The data quality problem got worse: Static B2B databases age faster than ever — job change rates accelerated post-2024. Starting with verified real-time data instead of stale lists isn’t optional anymore; it’s the baseline for deliverability.

What is a Clay API and why use it?

If you’re wondering, “Does Clay have an API?” the short answer is…Nope.

But don’t stop reading yet. You can still use Clay like an API with a clever workaround.

Here’s the idea: you create a table in Clay, fill it with the right columns to build a b2b contact database, then trigger updates or pull data via webhook or integrations. It’s not a traditional Clay enrichment API endpoint, but it behaves like one.

That means you can send it a lead’s email and get back enriched details (job title, LinkedIn profile, company info) – automatically.

Now, if you’re trying to do things like send data into Clay programmatically, wrap it with another tool, or do quick lookups, there are a few ways to make it happen:

Webhooks = best for sending data in and out

Every Clay table has a unique webhook URL. You can push data into a table from almost anywhere (a form submission, your CRM, or another app) and Clay will start processing it right away.

Once enrichment finishes, you can use HTTP actions to push the cleaned data back into your CRM, Google Sheet, or outreach tool.

It’s the most API-like workflow Clay offers, perfect for automating lead flow or enrichment jobs.

Wrap Clay in a third-party tool = best for light API proxying

If you absolutely need an endpoint, tools like Make, Zapier, or even a custom script on Replit can act as a middleman. They take in your API request, tell Clay to do its thing, and then return the results when ready.

Just remember: enrichment can take a minute or two, so you’ll need to add logic that waits for or checks results.

Enterprise-only People & Company API = best for quick lookups

If you’re on Clay’s Enterprise plan (which starts around $30,000/year), you can use their People and Company API. Send it an email or LinkedIn URL to get back basic profile info, or a company domain to get company details.

It’s great for lightweight lookups and simple enrichment, but doesn’t include deep data like verified emails, phone numbers, or revenue figures.

Now that you know how it works like a Clay data enrichment API, it’s time to put it to use. Let’s build a workflow that takes raw leads and turns them into ready-to-send prospects. But first…

The right starting point: feed Clay with live data, not static lists

Before we walk through the Clay workflow, there’s a step most guides skip entirely — and it’s the one that determines whether your whole campaign succeeds or fails.

The problem with static B2B databases: Most lead lists you buy or export from legacy tools were accurate when they were built. But B2B contact data decays at roughly 30% per year — meaning nearly a third of any static list is wrong by the time you use it. Wrong email = bounce. Bounces = deliverability damage. Deliverability damage = your entire domain’s sending reputation at risk.

Generect is built differently: it’s a real-time B2B lead search engine that pulls verified contacts on demand, instead of reselling records from static databases. Every contact is fetched and verified at the moment you request it — not six months ago when some vendor last refreshed their database.

The result: 98% email validity rate, verified at the point of export. Compare that to industry-average bounce rates of 8–15% from static list providers.

Generect is also selected for Google for Startups (1 of 35 companies globally), receiving $100K funding and $350K in Google Cloud credits — which funds the infrastructure for live, on-demand data at scale.

Here’s the practical delta this creates in a Clay workflow: when we run the same enrichment waterfall in Clay on Generect-sourced contacts vs. a purchased list export from a static database, the Generect batch skips the email-finding steps entirely (email already verified at export) — which cuts Clay Data Credit spend on that list by roughly 60–70%. You’re not paying Clay to find and verify emails that are already found and verified. That’s the economics of starting upstream.

Here’s the data comparison from running both source types through the identical Clay enrichment table:

Metric	Static database export (Apollo/ZoomInfo)	Generect real-time feed
Email validity on pull	72–85% (rest bounce or catch-all)	98% (verified at moment of request)
Clay email-finding credits consumed	~3–5 credits/lead (waterfall)	0 — already done
Time-to-usable-list	Export → Clay → verify → filter → 3–4 steps	API pull → Clay table → enrich signals — 1 step skipped
Data age	Last refreshed: unknown (often 6–18 months)	Fetched live at moment of query
Data Credit savings vs. static list	Baseline	~60–70% savings on email-finding step

The Generect → Clay workflow (step by step)

Here’s exactly how to use Generect as your live data source and Clay as your enrichment and personalization layer:

Step 1: Define your ICP in Generect and pull verified contacts

Use Generect’s API or UI to search for contacts matching your ICP — for example: SaaS companies, 50–500 employees, Head of Sales or VP Sales title, US-based. Generect returns verified contacts fetched in real time, not from a static database.

API call example (via Generect’s HTTP endpoint):

GET https://api.generect.com/search/contacts
  ?title=Head+of+Sales,VP+Sales
  &industry=SaaS
  &employee_count=50-500
  &country=US
  &limit=500

Response includes: first_name, last_name, email (verified), linkedin_url, company_name, company_domain, job_title, employee_count.

Step 2: Push the verified contacts into Clay via webhook

In Clay, create a new table and copy its webhook URL. Then POST your Generect results to that webhook — each contact becomes a row. Clay columns to pre-map:

Clay column	Generect field	Notes
First Name	first_name	Used in email personalization
Last Name	last_name	CRM sync
Verified Email	email	Already validated — skip re-verification step
LinkedIn URL	linkedin_url	Feed to Claygent for research
Company Domain	company_domain	Company enrichment trigger
Job Title	job_title	ICP scoring column
Employee Count	employee_count	Size filter
Data Source	“generect-live”	Static label so you can filter by source later

Step 3: Skip email-finding enrichment (already done)

This is where you save real Clay Data Credits. Since Generect already verified the email, you skip the entire email-finding waterfall in Clay (typically 3–5 enrichment providers, each consuming credits). Add a conditional rule: If “Verified Email” is not empty → skip email enrichment steps. On a list of 500 contacts, this alone saves roughly 1,500–2,500 Data Credits.

Step 4: Run Claygent for signal research on your best leads

Score your leads first (see next section), then run Claygent only on the top-tier accounts. A Claygent prompt that works well:

Visit {{LinkedIn URL}} and tell me:
1. Their most recent post topic (1 sentence)
2. Any job change in the last 90 days (yes/no + details)
3. A relevant company trigger (funding, hiring, product launch) from their company page

Return as JSON: {recent_post, job_change, company_trigger}

Claygent outputs land in new columns — Recent Post, Job Change Flag, Company Trigger — which feed your email personalization formula in Step 5.

Step 5: Write personalized first lines with Clay AI formula

In a new Clay column called Opening Line, use a formula like:

AI: Write a 1-sentence cold email opener for {{First Name}} at {{Company Name}}.
Context: {{Company Trigger}}. Their recent LinkedIn activity: {{Recent Post}}.
Keep it under 20 words. Be specific, not flattering.

Step 6: Push to your sending tool

Add an HTTP API action or native integration to push enriched rows to Smartlead, Instantly, or Reply.io. Map Verified Email, First Name, Opening Line, and any other personalization columns into your sequence template.

The end-to-end setup takes 2–3 hours the first time. After that, it runs automatically — pull fresh contacts from Generect, push to Clay, get personalized, launch campaign.

What this looks like in a real campaign (Marharyta’s workflow at Generect): A SaaS outbound team targeting 500 Head of Sales contacts in the US runs this exact workflow. Starting from Generect (real-time fetch, 98% valid emails), they skip Clay’s email waterfall entirely — saving ~1,500 Data Credits. They run Claygent on only the top 30% by ICP score (150 leads), costing ~300 Actions instead of 500. Total Clay credit spend: roughly 40% lower than a comparable workflow starting from a static Apollo export, while the email list is fresher and verified at the day of pull, not months prior. Clay-powered campaigns using this approach routinely hit 40–45%+ open rates, well above the 20–25% industry average for generic static-list outreach.

How do you set up a cold outreach workflow in Clay?

Setting up your workflow in Clay is like building the foundation of a house. If the base is solid, everything else (personalization, deliverability, and automation) works smoothly. Let’s walk through it from scratch.

1. Start with a smart table structure

Before you even pull in leads, decide what info you need. Your table is your command center.

At a minimum, add columns for:

Column name	Example data	Why it matters
First name	Sarah	For personalization in email and LinkedIn.
Last name	Lopez	Keeps records clean for syncing with CRMs.
Full name	Sarah Lopez	Needed for some enrichment APIs and email tools.
Company	GrowthFlow	Core personalization element.
Website domain	growthflow.io	Used for company enrichment and email finding.
Email	[email protected]	Direct contact method, used for outreach and verification.
Job title	Head of Marketing	Filters for ICP match, used in personal intros.
LinkedIn URL	linkedin.com/in/sarahlopez	Pulls social data, recent activity, and profile photo.

2. Grade your leads so you focus on the best

Not all leads are created equal. Clay lets you score them, so you spend time where it counts.

You can grade leads by:

• Job title match (e.g., only C-level or Directors)
• Company size
• Industry relevance
• Recent activity (like funding rounds or hiring)

Set up a scoring column with simple rules: higher score = hotter lead. This way, when you’re ready to send, you can filter and prioritize instantly.

3. Remove duplicates before they cause trouble

Nothing feels less personal than getting the same email twice. Clay’s deduplication step solves that.

You can:

• Deduplicate based on email address
• Or on company domain if you only want one contact per company

Pro tip: Run deduplication right before exporting to your sending tool. That way, no rogue duplicates slip through after enrichment.

Your workflow’s ready, but structure alone won’t win replies. The magic happens when you enrich those leads and filter out the noise, so you’re only talking to the right people.

How to enrich leads and build filters?

Enrichment is where your raw lead list turns into a goldmine. You’re not just finding names. You’re uncovering details that make your outreach personal and relevant. But if you’re not careful, you can burn through credits fast.

Let’s set it up the smart way.

1. Build enrichment recipes that work for you

In Clay, an “enrichment recipe” is simply the sequence of steps you run to add missing info.

Here’s how to make one:

1. Start with your must-haves. For example: verified email, job title, LinkedIn profile.
2. Choose your sources. Clay lets you pick from multiple data providers (150+ in 2026).
3. Stack them in order. If one provider can’t find the data, Clay moves to the next → this is your waterfall.

The waterfall approach is key. It means you only use credits from the next source if the first one fails, so you get maximum coverage without overspending.

Credit-saving tip for 2026: The single biggest Clay credit optimization is filtering before AI steps. Run Claygent and AI-formula columns only on rows where email is verified and intent signal is present. Teams that do this typically cut AI credit usage by 60–80% with no quality loss.

2. Use conditional logic to run only what’s needed

Not every lead needs every enrichment step. Conditional logic (the classic if–then rules) lets you skip unnecessary lookups.

For example:

• If the email column is already filled, then skip the “Find Email” step.
• If the company size is known, then skip the company enrichment.
• If the LinkedIn URL is missing, then run the LinkedIn search.

In Clay, you can set these conditions right inside your recipe. This keeps your workflow lean and ensures you’re only paying for what you actually need.

3. Combine recipes and filters for clean, ready-to-use lists

Once your enrichment steps are in place, add filters to remove bad or incomplete records:

• Filter out leads with missing emails.
• Filter out contacts without a job title that matches your target.
• Filter out unverified emails to protect deliverability.

This is the “polish” stage: your data is fresh, relevant, and ready to feed into personalization or sending tools.

With enrichment recipes, waterfall credit use, and smart if–then filters, you’re not just filling in blanks. You’re building a high-quality, cost-efficient lead list that’s tailored for your outreach goals.

Once your list is clean and powerful, you might want to connect Clay to tools it doesn’t integrate with directly. That’s where HTTP API actions open up a whole new world.

How to use Claygent for deep prospect research

Claygent is Clay’s built-in AI web agent — and in 2026 it’s become one of the most-used features in advanced Clay workflows. It’s not just an AI text generator; it’s an agent that actually browses the web on your behalf, reads pages, and returns structured answers at scale across your entire table.

Think of it as a research assistant who can visit 500 LinkedIn profiles or company websites overnight and fill in columns based on what it finds.

What Claygent can do

• Visit a company’s website and answer “What does this company actually sell?” — far more accurate than job-title guessing.
• Read a LinkedIn profile and surface recent posts, career changes, or stated priorities.
• Check a pricing page and extract the plan tier a prospect is most likely on.
• Scan a news/press page for recent company triggers (funding, launches, partnerships).

How to use Claygent without burning credits

Claygent is powerful but costs more per row than basic enrichment. The right approach:

1. Run basic enrichment first (email, title, company size) — cheap and fast.
2. Apply your ICP filter — keep only the rows that qualify.
3. Score leads and run Claygent only on leads above your score threshold.

Teams that follow this sequence typically run Claygent on 20–30% of the original list — getting the deep research where it matters, without paying for it on leads that won’t convert anyway.

Once your list is clean and powerful, you might want to connect Clay to tools it doesn’t integrate with directly. That’s where HTTP API actions open up a whole new world.

How do Clay HTTP API actions work?

Sometimes Clay’s built-in integrations aren’t enough. That’s where Clay HTTP API actions come in. They let you connect Clay to almost anything with an API: from your CRM to a custom meme generator.

If you can make a web request, you can make Clay talk to it.

An HTTP API action in Clay is simply a request you send out to another service. You tell Clay:

• Verb → what you want to do.
  • GET = fetch data
  • POST = send data
  • PUT/PATCH = update data
  • DELETE = remove data
• Clay API endpoints → the URL you’re sending the request to.
• Headers → extra info the service needs, like API keys or content type.
• Body → the data you’re sending (usually in JSON).

Clay makes this friendly by letting you use dynamic columns inside your request. That means if you have a “First Name” column, you can drop it straight into the API body without hard-coding it.

Let’s say you’ve enriched your leads in Clay and you want to call Generect’s API to pull additional verified contact data for leads that slipped through your initial ICP search:

1. Add an HTTP API Clay action step.
2. Choose GET as your verb.
3. Set the endpoint: https://api.generect.com/enrich/email?domain={{Company Domain}}
4. Add your Generect API key in the Authorization header.
5. Clay calls Generect for each row and stores the verified email in a new column.

You’re no longer limited to what Clay ships with → you can bolt on any tool, script, or service you can reach via an API.

You’ve learned how to make Clay talk to almost anything. Now let’s hook it up to your cold email tool so the enriched data flows straight into your campaigns.

How to connect Clay to cold email tools?

Once your leads are enriched and filtered in Clay, the next step is obvious: getting them into your sending tool. The good news? Clay makes this almost effortless.

If you’re using, for example, Reply, Clay has a native connector, yet you’ve got a few options:

Option A = Zapier

1. In Zapier, set Clay as the trigger app (new or updated row).
2. Choose Reply.io as the action app.

3. Map the fields so each new or updated Clay row becomes a new prospect in your sending tool.

Option B = HTTP API

If your email tool has an API (and Reply does), you can:

• Add an HTTP API action in your Clay table.
• Use a POST request to create a contact in your sending tool.
• Map Clay’s dynamic fields (like {{Email}} and {{First Name}}) into the request body.

Both methods keep your leads flowing automatically. No manual work required.

Option C = native integrations like Reply.io

Some cold email platforms, like Reply.io, now have a native Clay integration. No Zapier required.

With Reply’s integration, you can:

• Create contacts automatically from Clay (no manual CSV uploads).
• Push contacts directly into campaigns for instant outreach.
• Keep data in sync → if Clay updates a contact, Reply gets the update too.

How to set it up:

1. Get your Reply.io API key (Settings → API Key in Reply).
2. In Clay, add a new enrichment step, search for “Reply.io,” and pick Create Contact.
3. Map your columns → Email and First Name are required, but you can add extras like LinkedIn URL, City, or Custom Fields.
4. Run the column to create contacts in Reply.
5. Add another enrichment step: Push Contact to Campaign, choose your b2b saas lead generation sequence, and run it.

Now, any enriched lead in Clay can be live in a Reply campaign within seconds, ready for multi-channel outreach across email, LinkedIn, calls, SMS, and WhatsApp.

Which outreach tools pair well with Clay?

As you probably already know, to actually start conversations, you need a sending tool. The right one depends on whether you’re all-in on cold email or want multichannel sequences.

If you want to go beyond cold email, Reply.io is a strong partner.

• Channels = email, LinkedIn, calls, SMS, WhatsApp.
• Deliverability tools = inbox rotation, warm-up, and domain health tracking.
• AI SDR = helps automate prospecting and follow-ups.
• Verification = built-in email verification to cut bounces.
• Conditional campaigns = different follow-up paths based on prospect actions.

Tool	Best for	Key features when paired with Clay
Smartlead	High-volume cold email	AI personalization for subject lines & intros, multiple inbox sending, detailed campaign analytics
Woodpecker	Reliable, simple outreach	Personalization fields mapped from Clay, deliverability safeguards, automated follow-ups
SmartReach.io	Outreach at scale with safety	Email verification, sending limits, advanced scheduling, Zapier/API sync with Clay
Lemlist	Visual personalization	Personalized images/videos from Clay data, LinkedIn + email campaigns
Instantly	Fast deployment	Multiple inbox management, AI sending optimization, quick scaling from Clay leads
Outreach	Enterprise-level automation	Multichannel workflows, CRM syncing, detailed activity tracking
Salesblink	All-in-one prospecting & outreach	Built-in prospecting, email + LinkedIn automation, pipeline management
La Growth Machine	European multichannel focus	Automated LinkedIn + email combos, GDPR-conscious workflows

What are common pitfalls and how to avoid them?

Clay can automate a huge part of your outreach, but if you skip quality control, you risk wasted credits, poor deliverability, and low reply rates.

Here’s a quick-reference table of the 9 biggest mistakes and exactly how to fix them.

Pitfall	Why it’s a problem	How to avoid it
1. Over-automation without quality control	Bad or off-tone messages slip through unnoticed.	Test on a small batch first, manually review 5–10% of outputs, and check AI-generated lines for tone.
2. Ignoring filters and enriching the wrong leads	Wastes credits and risks emailing DNC or invalid addresses.	Apply DNC/verification filters before enrichment, run MX checks, remove irrelevant industries early.
3. Sending generic, copy-paste messages	Messages feel cold and get ignored.	Use AI formulas with profile-specific prompts, insert dynamic columns, keep intros short and relevant.
4. Not scoring or prioritizing leads	Time wasted on low-value prospects.	Add scoring columns (role, size, funding), focus on top 20–30%, auto-pause low-score leads.
5. Burning credits with inefficient enrichment	Credits drain fast without improving data quality.	Use if–then conditions to skip existing data, apply waterfall enrichment, review steps monthly.
6. Starting from a stale static list	High bounce rate kills domain reputation before you even start.	Pull contacts live from a real-time source like Generect (98% email validity) instead of buying list exports. Contacts fetched on demand are verified at the moment of request — not months old.
7. Overlooking deliverability prep	Emails never hit the inbox.	Warm up domains, limit sends per inbox, rotate inboxes, watch bounce rates.
8. Failing to integrate with sending tools	Manual exports/imports slow you down and cause sync errors.	Use native/API/Zapier integrations, map fields once, test with dummy contacts.
9. Not testing messaging variants	No data on what works best.	Create 2–3 variations, split test groups, track and scale the winner.

By avoiding these nine pitfalls, you keep your Clay-powered outreach sharp, relevant, and inbox-friendly.

Where to go next? Resources and templates

You’ve got the basics of using Clay for cold outreach. Now it’s time to level up. The fastest way to improve is by tapping into resources that show you exactly how top users build and scale their workflows.

Clay University & Documentation

Clay University is your crash course in mastering the platform. The short, focused lessons walk you through everything from building tables and creating enrichment recipes to integrating Clay with your favorite tools. Keep the official documentation bookmarked — it covers everything from API calls to setting up advanced filters.

Built-in templates

Clay has ready-made table templates for lead generation, enrichment, scoring, and outreach workflows.

• For lead generation for startups, start with a template instead of building from scratch.
• Swap in your own sources, filters, and AI prompts.
• Save your modified template so you can reuse it in future campaigns.

Claybooks

For even more inspiration, check out Claybooks — “recipes” shared by expert users, packed with real-world workflows for finding, qualifying, and personalizing leads. You can copy a Claybook straight into your account and tailor it to your niche.

Join the Slack community

Clay’s Slack community is an active hub where GTM operators swap workflow screenshots, Claygent prompts, and enrichment hacks. It’s also where you can get quick answers from people who’ve already solved the problem you’re facing.

Webhooks & automation samples

If you want Clay to connect with anything else in your stack, webhooks are your friend.

• Use the samples in Clay’s docs to send leads to CRMs, spreadsheets, or custom dashboards.
• Trigger follow-ups automatically when a lead’s data changes.

Once you’ve mastered the basics, these automation patterns will save you hours every week.

And if you want to take things even further, think about pairing Clay’s enrichment and workflow power with a live data source like Generect.

Starting with fresh, verified leads means Clay spends less time cleaning up bad records and more time adding the details that make your outreach stand out. Generect contacts are fetched in real time — not resold from stale databases — which is why our email validity rate is 98%. It’s a simple shift in how you feed Clay, but it can make the difference between a campaign that lands in the inbox and one that destroys your domain reputation before you’ve sent 100 emails.

The workflow in one sentence: Use Generect to pull 500 verified, live-fetched contacts → push via webhook to Clay → skip email-finding (already done, saves ~60–70% of Data Credits on that step) → run Claygent on top 30% by ICP score → write AI first-lines → push to Smartlead or Instantly. That’s the full loop, and it runs automatically after the first setup. If you’re starting from a static export, you’re doing the same workflow but paying more credits for worse data. That’s the only real difference — and it’s a significant one at scale.