Last month a Series B founder DM'd me a screenshot. His team had burned through $4,200 of Clay credits in six weeks. The output? A spreadsheet of 8,000 "enriched" contacts. About a third had no email. Another third had emails that bounced. The ICP filter caught maybe 1,200 real prospects.
He asked me one question. Is Clay broken or are we?
The answer is neither. Clay is doing exactly what it was told to do. The team just wired it up like a contact database when it is actually a routing engine. You point Clay at a list, and Clay decides which provider to ask for which piece of data, in what order, with what fallback. If the routing is dumb, the output is dumb. If the routing is smart, you get a clean dataset for a fraction of what a single enrichment tool would charge.
I have built Clay tables for nine GTM teams in the last two years. Here is the playbook I wish someone had handed me on day one.
What Clay actually is, and what it is not
Clay is not an enrichment provider. It is a spreadsheet that talks to 75+ providers and lets you stack them in sequence. The product calls this waterfall enrichment. The idea is simple. Ask the cheapest provider first. If it returns nothing, ask the next one. Keep going until you have the field or you run out of providers.
You pay a credit each time a provider returns a hit. So a contact that gets resolved on the first try costs one credit. A contact that takes four tries costs four. Over a list of 10,000 prospects, the order of your waterfall is the single biggest cost variable in the entire system.
People get this wrong because the Clay interface makes it look like a CRM. It is not. Treat it as a routing layer that sits between your CRM and your enrichment vendors, and the rest of the playbook makes sense.
Clay is a routing engine, not a contact database.
Every team that bleeds credits is using Clay like HubSpot. Every team that wins with Clay is using it like a Make.com workflow with 75 enrichment endpoints behind it.
The three waterfalls that matter
Most teams build one giant waterfall and stuff every field into it. That is the wrong shape. You want three separate waterfalls, run in this order.
1. The email waterfall
The expensive one. Email finding is hard because B2B emails are fragmented across four data sources that overlap inconsistently. A good email waterfall has 5 to 7 providers in sequence, starting cheap and ending expensive.
My default order, based on cost per valid email and hit rate on US/EU B2B contacts:
- Findymail (cheap, 50 to 70% hit rate on common ICPs)
- Datagma (cheap, fills different gaps from Findymail)
- Prospeo (mid-tier, strong on European data)
- Hunter.io (mid-tier, fills technical and dev personas well)
- ContactOut (expensive, last resort for senior titles)
After the waterfall, run a verification step through NeverBounce or MillionVerifier. Never push an unverified email into a sequence. The deliverability damage from one batch of bounces will outlast the campaign by months.
2. The phone waterfall
If you do not do cold calling, skip this. If you do, build a separate waterfall with three providers: Datagma, Cognism, and ContactOut. Mobile numbers are roughly 8x the cost of emails and the hit rate is half. Budget accordingly.
3. The company enrichment waterfall
The cheap one, but the one most teams skip. Before you even try to find a contact, you want to know if the company is worth enriching at all. Hit Clearbit (now HubSpot Breeze), then BuiltWith for tech signals, then your own custom ICP scoring step.
Run this waterfall first. It will kill 40 to 60% of your input list before you ever touch the expensive email providers. That is where the real savings live.
The order matters more than the providers
I have run the same 5,000-contact list through identical waterfalls with the providers reshuffled. The cost difference between the best and worst ordering was 4.1x. Same providers. Same hit rate. Different bill.
The rule is dumb but useful. Sort your waterfall by cost per hit, ascending. Cost per hit is not the same as cost per attempt. Cheap providers with low hit rates can be more expensive in practice than mid-tier providers with high hit rates, because Clay charges you on every successful return down the chain.
Two things make this hard to eyeball:
First, hit rates vary wildly by ICP. Findymail is great on US sales personas and weak on European engineering personas. Datagma is the opposite. Run a 500-contact pilot on a sample of your target list before you commit to an order.
Second, providers update their data weekly. The waterfall that worked in January 2026 may not be the best one today. Re-test every quarter.
The five mistakes that burn credits
After auditing a dozen Clay setups, the same five mistakes show up every time.
1. Running enrichment on the full list, not the filtered list
If your input list has 10,000 LinkedIn URLs and only 3,000 match your ICP, do not enrich all 10,000. Filter first, enrich second. Sounds obvious. Almost no one does it.
2. Using HTTP API columns instead of native integrations
Clay has native integrations for most providers. The native version is rate-limited, cached, and credit-billed correctly. The HTTP API version often double-charges because it bypasses Clay's cache. I have seen teams burn 30% extra credits because they copy-pasted a curl command instead of dragging in the integration.
3. No verification step
Unverified emails will tank your sender reputation in a week. Verification is one extra credit per email and saves your entire outbound program. Skip it once and you will spend three months rebuilding domain warmup. I have written about cold email deliverability in detail if you want the full picture.
4. Treating Clay as a permanent system of record
Clay tables are workspace, not storage. Once a record is enriched and verified, push it to HubSpot, Attio, or Salesforce. Then archive or delete the row in Clay. Teams that keep 500K rows in Clay are paying for nothing.
5. No cost ceiling per row
Clay lets you set a max-credit cap per row. Most teams leave it off. With it off, a single record can chew through 15 credits in a long waterfall. Cap it at 5 or 6. If a contact is that hard to find, they are probably not in your ICP anyway.
Wiring Clay into HubSpot or Attio
The integration story has gotten better in the last year, but it is still finicky. Here is what works.
Push, not pull
Do not let Clay continuously sync with HubSpot. The two-way sync feature looks clean and creates a mess inside a quarter. Pick one direction. Clay pushes new records to HubSpot when they are ready. HubSpot does not push anything back. Run any updates as one-off Clay imports when you need them.
Match on company domain, not company name
Company names are the most unreliable identifier in B2B data. "Acme Inc" and "Acme, Inc." and "Acme Corporation" will all create separate records if you match on name. Match on domain. If domain is missing, kick the record back to the company waterfall.
Use a staging property
In HubSpot, create a custom contact property called clay_status with values like "pending", "enriched", "verified", "exported". Push from Clay with status "enriched", then run a separate verification flow, then flip to "verified". Only "verified" records get into sequences. This one trick has saved more outbound programs than any other automation I have built.
Run the push through n8n if you can
Clay's native HubSpot push is fine for simple use cases. For anything custom, I move the export step into n8n. It gives you proper error handling, retry logic, and a queue you can pause when something looks off. Clay-to-CRM is the most common breakpoint in any RevOps stack. Treat it accordingly.
The cost model nobody publishes
Clay's pricing has changed three times since 2024. As of May 2026, the math looks roughly like this for a typical Series A or B B2B outbound program.
For a 5,000-contact monthly enrichment program with a 70% verified-email hit rate:
- Clay subscription: $349 to $800 per month depending on tier
- Provider credits: $400 to $900 per month (varies by waterfall design)
- Verification: $50 to $100 per month
- Total: $800 to $1,800 per month, all-in
That is roughly $0.16 to $0.36 per verified contact. Compare it to ZoomInfo at $25K to $50K per year for a similar volume and similar quality, and Clay wins on cost by 3 to 5x. It also wins on data freshness, because providers update independently and the waterfall reflects whatever is current.
Where Clay loses: it loses on tribal knowledge. A bad waterfall builder will spend twice as much for half the data of ZoomInfo. The savings are real but they are operator-dependent. That is the part most "Clay vs ZoomInfo" comparisons skip.
All-in monthly cost for a well-designed Clay setup running 5,000 verified contacts a month. Most teams overspend by 60 to 80% because the waterfall is wrong.
When Clay is the wrong tool
I want to be honest about this because every RevOps blog post on Clay reads like an ad.
Clay is wrong if you need a CRM. It is a workspace, not storage. Records there decay just like anywhere else.
Clay is wrong if you process under 500 contacts a month. The subscription cost outweighs the savings until you hit volume. Use a single provider direct.
Clay is wrong if you have nobody who can think in spreadsheets. The product is powerful because it gives you control of every column. If your team gets nervous around VLOOKUP, Clay will collapse on you in week three. Either hire an operator who lives in tables or pick a simpler tool.
Clay is wrong if your ICP is in a market the providers do not cover well. I have seen Clay struggle on Japanese mid-market and on European public sector data. Pilot before you commit.
What I would do this week if I were building from zero
If a founder asked me to set up Clay enrichment from scratch tomorrow, here is what I would do in order.
Day 1. Define the ICP in writing. Title patterns, company size band, geography, and at least two technographic or firmographic disqualifiers. No Clay table yet.
Day 2. Pull a 500-contact pilot list from LinkedIn Sales Nav. Drop it into a Clay table. Build only the company waterfall. Confirm 40 to 60% of the rows survive the ICP filter.
Day 3. Build the email waterfall on the survivors. Run it. Verify. Measure cost per verified contact and hit rate.
Day 4. Tweak the waterfall order based on which providers paid off and which were dead weight. Re-run on a second 500-contact pilot. Compare costs.
Day 5. Wire the push to HubSpot or Attio through n8n. Use a clay_status property so nothing goes live until verification clears.
Week 2. Scale to 2,000 contacts per week. Watch the cost per verified contact like a hawk. If it climbs above $0.35, something is wrong in the waterfall or the input list quality.
This is the boring version. It works. The exciting version is what most teams do, and what most teams have to throw away in month three.
Need a Clay table that actually pays for itself?
Book a 30-minute audit and we will rebuild your waterfall live. You keep the table. No retainer required.
Book an audit →FAQ
How much does Clay cost for a B2B startup?
Plan on $800 to $1,800 per month all-in for a Series A or B program enriching 3,000 to 5,000 contacts a month. The subscription itself is $349 to $800. The rest is provider credits and verification. If you spend less than that, you are probably under-enriching. If you spend more, the waterfall is misconfigured.
Is Clay better than ZoomInfo?
For under 50,000 contacts a year, Clay beats ZoomInfo on cost by 3 to 5x and matches or exceeds it on data freshness. Above that volume, ZoomInfo's flat-rate pricing starts to look reasonable. ZoomInfo also wins on intent data, which Clay does not natively provide. I have written a longer comparison in the CRM data enrichment guide.
How accurate is Clay's email finding?
Hit rate depends on the ICP, not on Clay. Across US B2B sales personas, a well-designed waterfall hits 65 to 80% verified emails. European technical personas drop to 40 to 55%. Asia-Pacific drops further. Always verify before exporting.
Can Clay replace a CRM?
No. Clay is a workspace, not a system of record. Keep your CRM as the source of truth and push enriched records from Clay into it. Teams that try to live in Clay end up with thousands of stale rows and no audit trail.
How long does it take to learn Clay?
A spreadsheet-comfortable operator can build a basic waterfall in a week. Getting good at column logic, formula columns, and conditional waterfalls takes about a month of daily use. Most teams underestimate the operator hours and end up with a setup nobody can maintain. Budget a real RevOps person, not a marketing intern.