Clay Credits and Actions: What Does an Enrichment Workflow Cost?

Clay pricing does not boil down to a simple per-contact or per-seat fee. A go-to-market enrichment pipeline consumes two distinct meters: platform execution (Actions) and third-party data or intelligence (Data Credits). Deciding whether Clay fits your budget requires breaking down the workflow into data waterfalls, AI web research, external provider keys, and downstream CRM sync.\n\nIf your team processes more than 500 target accounts per month and relies on multi-vendor contact discovery, Clay offers significant efficiency by charging only for successful data matches across vendor waterfalls. Conversely, if your monthly volume is under 200 leads or your CRM data is already verified, Clay's base subscription floor makes it uneconomic compared to native CRM automation or lower-cost workflow builders. For broader guidance across automation tools, see AI Workflow Automation Pricing Explained.\n\n## The dual-meter architecture: Actions vs. Data Credits\n\nClay separates the compute required to orchestrate a workflow from the variable cost of third-party data and language models. Understanding this split prevents unexpected bills.\n\n| Meter | What it measures | When it is billed | How to reduce consumption |\n| --- | --- | --- | --- |\n| Actions | Platform orchestration and execution: evaluating table formulas, running conditional logic, executing webhook triggers, normalizing text, and syncing records downstream to a CRM. | Charged per row or operation when an enrichment, formula, or integration step runs. | Batch operations, filter out low-intent records before running formulas, and use view-level auto-update controls. |\n| Data Credits | Marketplace data and AI consumption: acquiring verified emails, company technographics, firmographics, mobile numbers, or generating summaries with built-in AI models. | Charged upon successful data acquisition or per model prompt/token block. | Connect your own API keys (BYOK) for LLMs and data providers; target waterfalls to high-confidence providers first. |\n\nEvery AI prompt executed natively through Clay counts as one platform Action, while the model execution itself draws Data Credits. If an enrichment step queries a vendor and returns no valid data, Clay does not charge Data Credits or Actions for that failed lookup.\n\n### Bring Your Own Key (BYOK) economics\n\nOne of Clay's most effective cost-governance mechanisms is Bring Your Own Key (BYOK). Teams can connect their own API accounts for language models (such as OpenAI, Anthropic, or Perplexity) and data vendors (such as Apollo, Dropcontact, or Hunter).\n\nWhen BYOK is enabled for a specific step:\n1. Clay waives its Data Credit charge for that enrichment or AI task.\n2. The user pays the underlying provider directly at raw API rates.\n3. Clay continues to charge standard platform Actions for orchestrating the request and structuring the output.\n\nFor teams running high-volume AI research or teams that already maintain enterprise data provider contracts, BYOK dramatically reduces overall Clay spend. For an analysis of subscription allowances versus direct API meters, consult AI Subscription vs API Pricing.\n\n## How multi-provider waterfalls and Claygent consume credits\n\nEnrichment costs vary widely depending on whether a workflow queries basic corporate registries, multi-step email waterfalls, or autonomous AI agents.\n\n### Waterfall credit mechanics\n\nRather than relying on a single data provider that might have a 40% to 60% match rate, Clay allows operators to stack multiple providers in sequence (a waterfall). For example, a waterfall might query Provider A, then Provider B, then Provider C until a verified work email is discovered.\n\nClay's waterfall billing operates on a success-only model:\n- Zero charge for misses: If Providers A, B, and C fail to find a valid contact, zero Data Credits are deducted for those failed attempts.\n- One credit per standard match: When Provider D finds a verified email, Clay charges exactly 1 standard Data Credit for the valid result.\n- Premium data sources: Direct phone numbers, validated mobile dials, and specialized compliance datasets consume 2 to 5 Data Credits per successful match.\n\nThis structure eliminates the financial penalty of building deep waterfalls to maximize contact coverage.\n\n### Claygent AI web research\n\nClaygent is Clay's native AI web scraping and research agent. Instead of querying static databases, Claygent navigates live websites, reviews LinkedIn profiles, extracts executive biographies, or checks whether an account uses specific software.\n\nClaygent consumes Data Credits dynamically based on research depth:\n- Standard prompt research: Typically consumes 1 to 2 Data Credits per record for straightforward extraction (e.g., finding a company's pricing model or headquarter location).\n- Multi-step reasoning tasks: In-depth executive briefings or multi-page scraping tasks consume 2 to 4 Data Credits.\n- Variable pricing and reconciliation: For token-intensive reasoning models, Clay applies a withhold-and-reconcile mechanism. The platform reserves estimated credits upfront before the agent run and refunds unused credits immediately upon task completion with zero token markup.\n\n## Subscription tiers, overage rates, and rollover rules\n\nClay packages platform capacity into monthly or annual subscriptions. Buyers evaluating the platform will encounter both modern workspace tiers and classic legacy plans.\n\n| Plan tier | Monthly price (billed monthly) | Annual equivalent (billed monthly) | Included monthly allowance | Additional credit overage rate | Core capabilities & boundaries |\n| --- | --- | --- | --- | --- | --- |\n| Free | $0 | $0 | 500 Actions; 100 Data Credits | N/A | Unlimited seats and tables; 200 rows per table; BYOK supported. |\n| Starter (Legacy) | $149 | ~$149 | 2,000 credits/mo | ~$0.10 per credit | Basic enrichment; entry-level outbound prospecting. |\n| Launch (Modern) | $185 | $167 | 15,000 Actions; 2,500 Data Credits | Tiered top-up packs | Phone enrichment; Audiences search; reusable Functions; up to 50,000 rows/table. |\n| Explorer (Legacy) | $349 | ~$349 | 10,000 credits/mo | ~$0.08 per credit | Expanded waterfall volume for growing outbound teams. |\n| Growth (Modern) | $495 | $446 | 40,000 Actions; 6,000 Data Credits | Tiered top-up packs | Native CRM and data warehouse auto-sync; HTTP API integrations; webhooks; ad audiences. |\n| Pro (Legacy) | $800 | ~$800 | 50,000 credits/mo | ~$0.05 per credit | 20 concurrent workflow executions; priority data queues. |\n| Enterprise | Custom | Custom (typically $24,000+/yr) | Custom pools (100k+ credits/mo) | Volume-discounted blocks | SAML SSO; role-based access control (RBAC); dedicated customer success; custom sub-processor agreements. |\n\n### Rollover and overage policies\n\n- Subscription credit expiration: Included monthly subscription credits expire at the end of each billing cycle without rollover. Unused allowances do not carry over to the next month on self-serve plans.\n- Prepaid top-up blocks: If a workflow exhausts its monthly allowance, accounts can purchase prepaid credit top-up blocks. These overage credits scale from ~$0.10 per credit on lower tiers down to ~$0.05 per credit on Pro and Enterprise tiers. Top-up credit blocks retain rollover flexibility based on tier terms.\n- Unlimited team seats: Clay does not bill on a per-seat model for standard workspace members. Multiple SDRs, RevOps managers, and marketers can collaborate within one workspace without increasing the base subscription price. However, all users consume from the same shared pool of Actions and Data Credits.\n\n## Illustrative lead-enrichment ledger: 1,000 vs. 5,000 accounts\n\nBecause credit consumption depends on waterfall match rates and AI research depth, there is no universal \"cost per lead.\" To model realistic operating budgets, compare two representative outbound prospecting workloads using documented platform rules.\n\n### Scenario A: 1,000 target accounts (Firmographic triage & executive briefing)\n\nWorkload assumptions:**\n- 1,000 target company domains imported.\n- Firmographic & tech stack enrichment: 1 credit per account (100% resolution) = 1,000 credits.\n- Verified work email waterfall: 80% match rate across providers = 800 verified contacts (800 x 1 credit = 800 credits; 200 misses = 0 credits).\n- Claygent AI executive briefing (summarizing recent company initiatives for the VP of Sales): 2 credits per verified contact = 1,600 credits.\n- CRM export: 800 enriched records synced to CRM (800 Actions, 0 Data Credits).\n\nTotal consumption:\n- **Data Credits: 1,000 + 800 + 1,600 = 3,400 Data Credits.\n- Platform Actions: ~3,600 Actions (evaluating formulas, scraping steps, and CRM pushes).\n\nPlan fit and operating cost:**\n- On legacy tiers, this workload requires an **Explorer plan ($349/month) or a Starter plan ($149 base + 1,400 overage credits at $0.10 = $289/month).\n- Effective total cost ranges from $289 to $349 per month, yielding an illustrative cost of $0.29 to $0.35 per fully qualified lead.\n- On modern plans, 3,400 Data Credits fits the Launch plan ($185/mo) with a small top-up pack for the remaining 900 Data Credits.\n\n### Scenario B: 5,000 target accounts (High-volume pipeline generation)\n\nWorkload assumptions:**\n- 5,000 company domains imported.\n- Firmographic & tech stack enrichment: 1 credit x 5,000 accounts = 5,000 credits.\n- Verified work email waterfall (80% match rate): 4,000 verified contacts x 1 credit = 4,000 credits (1,000 misses = 0 credits).\n- Claygent AI executive briefing: 4,000 contacts x 2 credits = 8,000 credits.\n- Downstream CRM export: 4,000 records synced (4,000 Actions).\n\nTotal consumption:\n- **Data Credits: 5,000 + 4,000 + 8,000 = 17,000 Data Credits.\n- Platform Actions: ~17,000 Actions.\n\nPlan fit and operating cost:**\n- On legacy tiers, this workload comfortably fits the **Pro plan ($800/month, includes 50,000 credits).\n- Effective cost is $800 per month, or approximately $0.16 per fully qualified record when utilizing the plan's capacity.\n- The BYOK optimization: If the team connects their own OpenAI API key for the Claygent executive briefings, direct LLM API costs drop to ~$40 (assuming ~$0.01 per summary via gpt-4o-mini), while Clay Data Credit consumption falls from 17,000 to 9,000 credits. This allows the team to operate on Explorer ($349/mo) plus a direct LLM bill, saving over $400 per month.\n\n## Downstream costs: CRM sync and the complete GTM stack\n\nClay is an enrichment and orchestration engine, not a complete standalone CRM. When forecasting an automation budget, account for the entire commercial stack:\n\n1. CRM storage and user licenses: Pushing thousands of newly enriched accounts into platforms like HubSpot can trigger contact-tier upgrades or require paid sales seats. For CRM workflow boundaries, see HubSpot Starter vs Professional Workflows.\n2. Sales engagement sequencers: Sending outbound email campaigns requires external mailboxes, warming tools, and sending platforms (e.g., Smartlead, Instantly, or Outreach).\n3. Dedicated workflow backbones: For complex, cross-departmental operations involving billing, internal databases, and Slack routing, pairing Clay with a general-purpose automation platform like Make or n8n often proves more cost-effective than using Clay for non-GTM logic. Review AI Workflow Automation Platforms Compared for architectural trade-offs.\n\n## Contraindications: Who should NOT buy Clay\n\nClay is a specialized, premium GTM tool. It is an active anti-fit for several common business profiles:\n\n- Low-volume outbound teams (<200 leads per month): If your sales team only contacts 100 to 150 prospects monthly, Clay's $149/month or $185/month entry tier results in an effective cost exceeding $0.75 to $1.00 per lead before paying for sales engagement tools. Manual prospecting or entry-level contact databases like Apollo are far more economical.\n- Teams with clean, existing CRM data: If your database already contains accurate, verified contacts and your primary need is moving deal stages, assigning owners, and triggering task alerts, you do not need Clay. Use native CRM builders like HubSpot or general automation builders like Make.\n- General-purpose IT and internal operations: Clay's table-based canvas and credit model are optimized for prospect and company data. Using Clay to sync Stripe payments to accounting software or pipe customer support tickets to Slack is inefficient and expensive compared to n8n or Make.\n- Teams seeking simple no-code app integration: If your workflows require connecting general SaaS apps without data enrichment or AI research, choose a dedicated integration tool. See How to Compare AI Tool Pricing Units to evaluate alternative unit meters.\n\n## Buyer decision checklist before purchasing\n\nBefore entering a paid Clay subscription, run through these four practical verification steps:\n\n1. Measure your monthly volume: Count the unique domains and contacts your outbound team can realistically engage each month. Do not purchase a high-capacity tier based on aspirational lead lists that your sales reps cannot work.\n2. Test your match rate on the Free plan: Run a sample batch of 100 target accounts on Clay's free tier. Observe your actual waterfall email match rate and phone resolution rate to calculate an accurate credit multiplier.\n3. Audit BYOK viability: Check whether your engineering or RevOps team can provide company API keys for OpenAI, Anthropic, or Apollo. If yes, plan on using BYOK to bypass Clay Data Credit markups on high-volume steps.\n4. Compare against CRM-native enrichment: If your team already uses HubSpot Professional or Enterprise, evaluate whether native Breeze Intelligence credits cover your enrichment needs before introducing Clay as an external data silo. For a direct head-to-head comparison, consult Clay vs HubSpot.