Clay

What it is

Clay is a GTM data platform that looks like a spreadsheet, behaves like Zapier, and connects to every data broker in the B2B universe. Launched in 2017, it sat quiet for years as a niche tool for RevOps power users before exploding into the default "how modern outbound teams actually work" platform sometime around 2023. Today it is one of the most talked-about tools in the B2B GTM stack and the reference implementation for the "signal-based outbound" movement.

The core metaphor is a table. Each row is a person or a company; each column is either a piece of static data you uploaded, an enrichment column that calls an external provider, an AI column that runs Claygent (Clay's research agent) against the web, or a logic column that transforms the rest. You build a workflow by stacking columns: start with a list of domains, enrich to find decision-makers, run Claygent to research recent initiatives, score the fit with AI, filter, then push the qualified rows to your CRM or sequencer. It is the closest thing in 2026 to "Airtable for data-ops teams, with a hundred enrichment providers pre-wired in."

The feature that sets Clay apart from every other data tool is the enrichment waterfall. Instead of picking one provider and living with its coverage gaps, you chain them: try provider A first, fall back to B if A returns nothing, then C, then D. Clay only charges for successful lookups within the waterfall, and stops as soon as one provider finds the answer. For email finding, phone numbers, firmographic data, technographic signals — anything where no single provider covers the whole universe — the waterfall pattern is a 3–10× cost saver over running one provider flat, and the coverage is dramatically better. It is the single reason serious outbound teams rearchitect around Clay.

Category-wise, Clay sits in a weird spot. It tags itself as a GTM platform; we tag it as a CRM adjacent workbench. It is not a CRM — there is no pipeline UI, no deal-stage tracking, no seat-based team-selling surface. You still need HubSpot or Salesforce underneath. What Clay replaces is the horrible patchwork of Apollo-plus-ZoomInfo- plus-Clearbit-plus-a-Zapier-rat-king that most outbound teams assemble before they find Clay. That consolidation, plus the Claygent AI agent on top, is the whole value proposition.

Claygent deserves its own sentence. It is an AI research agent that runs against the public web (LinkedIn, company sites, news, job boards, earnings calls) and returns a structured field to your table. "Does this company mention generative AI on their careers page in the last 90 days?" is a column. "Which of their recent LinkedIn posts suggests an outbound-relevant pain?" is a column. It is the feature that makes Clay feel less like a spreadsheet and more like a research team you rent by the credit.

What we tested

In our testing across client engagements and internal work, we have run Clay in production for more than eighteen months across the Starter, Explorer, and Pro tiers. We have built lead lists from scratch, migrated clients off Apollo-plus-Clearbit stacks onto Clay-plus-a-CRM, set up signal-based outbound programs that trigger on job changes and funding announcements, and shipped the kind of "research 500 accounts in an evening" workflows that Clay exists for.

On the enrichment side we have tested waterfalls across every major provider — Apollo, ZoomInfo, Clearbit (now HubSpot Breeze), Lusha, Hunter, Findymail, Prospeo, Datagma, RocketReach, LinkedIn Sales Nav scraping, plus the cell-phone waterfalls that are Clay's strongest and most expensive feature. We have watched credit consumption patterns across tens of thousands of rows, tuned waterfall ordering to minimize cost, and built the provider-chain playbook for client engagements.

On the automation side we have connected Clay to HubSpot and Salesforce in both directions, pushed enriched rows into Instantly and Smartlead for sending, wired Clay to Slack for real-time lead alerts, and tested the webhook + API surface for custom integrations. We have run Claygent against hundreds of research tasks — recent funding, tech-stack detection, hiring signals, recent content, competitive intelligence — and calibrated when it is worth the credits versus when a cheaper provider would answer the same question.

None of what follows is a formal benchmark. Clay is the kind of tool where the "benchmark" is your own workflow: what does it cost to qualify a thousand accounts, how clean is the data on the other side, how much SDR time does it save. What we can offer is the texture of running Clay in production across real client budgets, the pricing surprises that trip up every new team, and an honest read on where the learning curve is worth it versus where a simpler tool would land faster.

What's good

The single biggest reason to use Clay is the enrichment waterfall. No other tool in the B2B data category gets this right. If you have ever paid $15k/year for a ZoomInfo seat and then watched 30% of the emails bounce, the experience of setting up a four-provider waterfall in Clay and watching coverage jump to 85%+ at a fraction of the cost is genuinely transformative. The math is simple: providers have different coverage sets, and stacking them lets you combine coverage without paying for redundancy. Clay is the only tool that implements this as a first-class primitive.

100+ data sources in a single interface is the second thing that makes Clay indispensable. Apollo, ZoomInfo, Clearbit, Lusha, Hunter, Findymail, Prospeo, Datagma, Ocean.io, LinkedIn Sales Nav, BuiltWith, Semrush, Crunchbase, PredictLeads — dozens more, all exposed as native columns. You no longer need a contract with each provider; Clay resells access and abstracts the API. For a small team, this alone replaces six vendor relationships, six onboarding calls, six seat licenses, and six places where data lives. The consolidation is the point.

Claygent is the AI research feature nobody else has matched. You can write a column prompt like "visit this company's website and their last three LinkedIn posts; summarize whether they're hiring for ML roles and what pain points they seem to have around data" and get a structured result per row. At scale — hundreds or thousands of rows — this is the kind of work that used to require a junior SDR or a contractor team. Claygent does it in minutes. It is not magic; the outputs need curation and the credits add up quickly. But the leverage is real.

The templates library is Clay's underrated onboarding asset. Hundreds of pre-built workflows — "enrich from Apollo search," "find decision-makers at companies that recently raised," "research accounts for account-based outbound" — let new users clone a known-working table and customize rather than build from scratch. The community around these templates is one of the most active in B2B SaaS right now, and the quality has compounded. Clay University (the docs and training site) is genuinely good, which is not something we say often about SaaS onboarding.

Where Clay earns its keep

• Enrichment waterfalls save 3–10× vs. single-provider enrichment, and coverage is dramatically better.
• 100+ native data sources collapse six vendor contracts into one subscription.
• Claygent scales "research this account" work from hours-per-SDR to seconds-per-row.
• Signal-based prospecting templates (job changes, funding, hiring, tech-stack) out of the box.
• Community templates library is the fastest path from zero to working outbound workflow.
• Power-user community on Slack and LinkedIn is one of the most active in B2B SaaS — answers are real.

Clay is not a CRM. It is the workbench that feeds your CRM — the place where raw lists become qualified, enriched, researched rows. Expect to run HubSpot or Salesforce underneath, and accept that Clay is the engine in front of it.

Signal-based prospecting is the workflow pattern Clay made mainstream. Templates for "people who just changed jobs into target titles," "companies that just raised a Series B," "companies hiring for specific roles that imply a buying signal" — these used to be custom data-engineering projects and are now half-hour table clones. For any outbound team that has moved past spray-and-pray, this is the real reason Clay shows up in every modern outbound stack.

Common pitfalls

Six predictable failure modes show up in almost every Clay engagement we advise on. Every single one of them is avoidable with thirty minutes of planning, and every single one of them costs real money when skipped.

Not understanding credit economics before running your first workflow. This is the #1 pitfall and it is expensive. A workflow that looks fine on paper — enrich 5,000 accounts, find the CMO at each, find the mobile phone, research with Claygent — can consume 500,000+ credits on a single run. That is five months of Explorer tier in one afternoon. Before you build, estimate: how many rows × how many credit-consuming columns × average credits per column. Then run it on 10 rows first and multiply. The Clay docs have a credit estimator; use it. Teams who skip this step routinely overshoot their monthly budget by 5–10× in their first real run.

Skipping enrichment waterfalls and wasting credits on single-provider lookups. New users often wire up one provider — usually Apollo, because they already know it — and treat Clay as a nicer Apollo UI. That is a 3–10× overspend versus a proper waterfall, and the coverage is worse too. The correct pattern is always a chain: start with the cheapest provider with decent coverage, fall back to progressively more expensive providers only when the cheap ones fail. Clay only charges for successful waterfall legs; the first provider that finds the answer wins. If you are not running waterfalls on every high-volume column, you are using Clay at beginner mode.

Treating Clay as a primary CRM. Clay is not a CRM. It has no pipeline, no deal stages, no team-selling surface, no forecasting, no commission tracking, no meeting scheduler. It is a data workbench. Teams who try to run deals out of Clay tables end up rebuilding half a CRM badly. The correct pattern is Clay for data enrichment and list-building, synced into HubSpot or Salesforce for pipeline and deal management. Keep the boundary clean and both tools work together. Blur it and you will be miserable.

Over-engineering when templates would work. Clay's community templates library covers most standard outbound workflows. New users often ignore it and spend a week building a workflow from scratch that already exists, better, in the template library. Check the library first; clone and modify before you build fresh. The templates encode real operator knowledge about which providers to chain, which Claygent prompts actually work, and which signals to filter on.

Ignoring the community templates library entirely. Related but worth calling out separately: Clay's community templates — shared by actual RevOps operators — are where the state of the art in signal-based outbound actually lives. If you are building waterfalls from scratch without checking what the community already solved, you are losing time. The templates are free, cloneable, and usually better than what you would build on your own in the first month.

Not connecting Clay to downstream CRM or sequencer. Clay generates enriched, qualified, researched rows. If those rows stay in Clay tables, the work has no output. Wire Clay to HubSpot / Salesforce for CRM sync, to Instantly / Smartlead / Outreach for sending, and to Slack for real-time alerts on high-signal rows. Clay on its own is a research tool; Clay wired into your GTM motion is the leverage the product is designed for.

What's not

Clay is not a CRM and you have to internalize that before you buy. There is no pipeline view, no deal-stage tracking, no commission reporting, no team-selling surface, no forecasting. Every Clay deployment we have seen sits alongside HubSpot or Salesforce, and the division of labor is clean: Clay builds and qualifies the list, the CRM runs the deals. Teams that try to flatten this distinction end up rebuilding half a CRM inside Clay tables, badly. Accept the architecture Clay implies.

The learning curve is real. Clay is a power-user tool that assumes you think about data like an engineer — waterfalls, branching logic, rate limits, credit budgets, API pagination. First-time users routinely spend two weeks before the product "clicks" and another two before they are producing real output. Teams that expect a five-minute onboarding end up bouncing off. Budget for the ramp; it is worth it if you make it through, but it is not optional.

Credit pricing is the second structural friction. Credits are consumed at wildly variable rates per action, the cost model is non-linear in ways that defeat spreadsheet intuition, and the same workflow run on different data can consume 5× different credit totals depending on which waterfall leg fires. Every RevOps team we have advised on has had at least one "how did we burn through a month of credits in a week" conversation in their first quarter. The March 2026 pricing split (Data Credits vs. Actions) is an attempt to clean this up — it helps — but the underlying variability is inherent to running enrichment waterfalls against third-party providers.

The UI and mobile experience trail modern CRMs by a visible margin. Clay is built for desktop power users running big tables in landscape browser windows. On mobile it is barely usable, and the in-table editing experience is functional rather than delightful. For the people who need Clay, this is a non-issue — the work happens at a laptop with a coffee. But reps who spend their day in a phone-first workflow will find the product alien.

Who should use it

If you are a RevOps team at a B2B SaaS company running serious outbound, Clay is close to mandatory in 2026. The enrichment waterfalls alone will pay for the subscription in vendor-contract savings. The Claygent research layer will add leverage no SDR team can match manually. The templates library will shorten time-to-first-working-workflow from weeks to days. Explorer at $349 is the starting tier; most serious RevOps teams graduate to Pro at $800 within the first quarter as they build more workflows.

For founder-led B2B sales — a founder or early-stage team doing outbound themselves — Clay is one of the highest-leverage tools you can put in your stack. The templates library means you do not need a data engineer to get working workflows running; the waterfalls mean you do not need six separate data contracts; Claygent means you can research accounts at a scale no early-stage team could hire for. Start on Starter ($149) with the 14-day unlimited trial, work through two or three templates during the trial to learn the product, then step up to Explorer when the trial ends.

For outbound ops teams inside larger organizations, Clay is the tool that consolidates your patchwork of Apollo + ZoomInfo + Clearbit + Zapier into a single workflow surface. The political win is real — one subscription, one training curve, one point of ownership for "where does the lead list come from." The credit economics will need attention, but the alternative is living with four separate data contracts and the integration tax that comes with each. Pro ($800) or Enterprise is the right tier here.

For data-first SDR organizations that compete on the quality and timing of their outbound, Clay is the reference implementation of signal-based prospecting. Job changes into target titles, recent funding rounds, hiring signals, tech-stack changes, competitive intel — all of it expressible as Clay columns, all of it routable to HubSpot or directly to Instantly / Smartlead. If your SDR strategy is "reach buyers at the moment they have a need," Clay is the tool that operationalizes it. Nothing else in the category is close.

For agencies doing research-heavy outbound for clients — lead-gen agencies, outbound-as-a-service shops, embedded RevOps consultants — Clay is the service delivery platform. The templates library becomes your service catalog; Claygent becomes your research bench; the waterfalls become your margin. Agencies running Clay can deliver lead lists at a quality and cost that single-provider competitors cannot match. Pro ($800) or Enterprise is the right tier; the ROI math is obvious within the first client engagement.

For SMBs who want a simple CRM, Clay is the wrong tool. Go use HubSpot's free tier, or Pipedrive, or Attio. Clay assumes you have outbound volume and data-ops capacity that most SMBs do not, and the credit economics punish teams who do not have the workflow to justify them. If you have to ask whether Clay is for you, it probably is not — yet.

Verdict

Clay is the most important new tool in the B2B GTM stack in 2026, and also the one most likely to blindside a new buyer with its bill. The enrichment waterfalls are genuinely transformative; the Claygent AI research agent adds a layer of leverage that no competitor has matched; the templates library and community make the learning curve survivable. For the RevOps team, the founder-led seller, the outbound ops lead, and the research-heavy agency, Clay is the tool that rearchitects the GTM motion around data-first workflows, and the rest of the category is chasing.

We rate it 8.6 / 10. It loses points for the credit-economics confusion that trips up every new user, for the learning curve that excludes teams without data-ops capacity, and for the opaque Enterprise pricing. It gains them for being the category-defining implementation of enrichment waterfalls, for Claygent, for the community, and for the sheer leverage the tool delivers once you are past the ramp.

If you are on the fence, take the 14-day unlimited trial on Starter and build two workflows from the templates library. Run them. Watch the credit consumption. If the output maps to your outbound motion and the credit math pencils out, Explorer at $349 is the right next step. If it does not, you have saved yourself a learning curve. Either way, you will know.