How to use Clay for outbound, and when to skip it

Sunny Ray · Oct 22, 2026 · 7 min read

To use Clay for outbound, import or pull a list of target accounts and people, run enrichment columns to add the data you need (verified emails, company facts, recent signals), use an AI column to write one specific line per prospect from that data, then push the finished rows to a sending tool like Instantly or Lemlist. Clay doesn't send at volume itself. It's the research and enrichment layer that sits between your list and your sender.

The more useful question is whether you need it at all. Most Clay tutorials assume every campaign needs heavy enrichment. Ours doesn't, by default. We reach for Clay when the offer needs a reason to write to one specific person, and skip it when the offer itself is the reason. Below are both paths.

What Clay is good at

Clay is a spreadsheet where each column can call a data provider, a scraper or an AI model. Its strength is combining sources: if one provider doesn't have an email, the next column tries another. It can pull signals like hiring, funding or tech stack, and it can turn all of that into a personalized line.

It's become a standard tool for the people who build outbound systems. In RevGuild's 2026 market report, Clay was named in 150 of 303 GTM engineer job postings from June 2025 to June 2026. If you hire someone to run outbound, there's a good chance they'll ask for it.

Our default without Clay

On a team call on August 27 I described our normal workflow, and Clay wasn't in it. Deep enrichment means running the list through Clay, where you can go crazy with enrichment. But that's not what we usually do.

Our default is three steps:

1. Apollo for the list. Filter to the titles, industries and company sizes that match the ideal customer.
2. MillionVerifier to check every address. On that same call we checked bounce rates across our campaigns and they were effectively zero, which we put down to verifying before sending.
3. Instantly to send, from warmed mailboxes on separate domains.

Why skip enrichment? Because our offer doesn't need it. We invite people onto a show. The invitation works because it's about them as a category: a founder in this market, an investor at this stage. A one-line reference to their latest funding round doesn't change much when the ask is "come talk about your work." The time and credits go further on a bigger, cleaner list.

When enrichment pays

Enrichment pays when the reason to write depends on something specific about the person or company. Some examples:

• Trigger-based outreach. You only want companies that just raised, just hired a head of sales, or just launched in a new market.
• Small, high-value lists. Two hundred enterprise accounts where each deal is worth a lot. A specific first line is worth the cost per row.
• Technical fit. Your product only works with a certain tool, and you need to know who uses it.
• Missing data. Your list source doesn't have emails for a niche, and you need a waterfall of providers to find them.

A consultant I spoke with on September 12 ran a stack built around exactly that kind of care: Clay, with Lemlist and Instantly for sending. But the part I remember is that a human reviewed 10 curated emails a day. Clay did the research. A person made sure each email was worth sending. For high-value, low-volume outreach, that's the right shape.

A simple Clay outbound table

If you're building your first Clay table for outbound, keep it simple. One table, a handful of columns:

1. Company domain. Your starting list.
2. Company enrichment. Size, industry, location. Filter out anything that doesn't fit.
3. Find people. The titles you sell to, at each company.
4. Email waterfall. Try providers in order until one finds a valid address.
5. Verification. Drop anything that doesn't verify.
6. One signal. The single fact that gives you a reason to write. Recent funding, a job post, a launch.
7. AI first line. A prompt that turns the signal into one natural sentence. Keep it short and check a sample by hand.
8. Push to sender. Send the finished rows to your sending tool's campaign.

Resist adding twenty more columns. Every column costs credits, and most of them won't change what you write.

Pairing Clay with a sending tool

Clay prepares rows. A sending tool handles the part that can break your domains: mailboxes, warm-up, daily limits, rotation and the reply inbox. Keep those jobs separate.

• Send from separate, warmed domains, never your main one.
• Keep daily volume per mailbox low.
• Verify before every push, even if the data looks fresh.
• Map your AI first line to a custom variable in the sender, so you can see exactly what each person receives.
• Leave links out of the first email.

Do the mailbox math before you enrich

Here's a check that saves a lot of Clay credits: work out how many emails you can actually send per day before you decide how many rows to enrich.

Sending capacity is set by your mailboxes, not your table. On our August 27 team call we set up a new batch at three mailboxes per domain to start, at about 10 emails per mailbox per day, which works out to roughly 30 emails a day per domain. Charles's guideline is to grow to about 20 mailboxes per domain for podcast invitation campaigns, and about 10 per domain for direct sales campaigns, which carry more spam risk. Even then, each mailbox sends only a handful of emails a day.

So if your setup can send a few hundred emails a day, enriching 20,000 rows this week is paying for data that will be stale by the time you reach it. Enrich in batches that match your sending capacity, a week or two ahead at most. Fresh data also bounces less.

The same math tells you when enrichment is worth more. If you can only send 100 emails a day to a small, high-value market, each one should be as good as you can make it. If you're sending thousands of invitations, a clean list beats a clever line.

Costs to watch

Clay runs on credits, and enrichment burns them quickly. The costs that sneak up on teams:

• Waterfalls on bad lists. Running five providers on rows that don't fit your profile. Filter first, enrich second.
• AI columns on every row. Write the first line only for rows that passed verification.
• Re-running tables. Every refresh costs credits. Decide how fresh the data really needs to be.
• Signals you don't use. If a column doesn't change the email, delete it.

A simple check: compare cost per enriched row with what a better first line actually does to your reply rate. If you can't see a difference after a few thousand sends, go back to the simple path.

The short version

Use Clay when the reason to write lives in the data. Skip it when the reason to write lives in the offer. And whichever path you take, verify every address and protect your sending domains.

If you'd rather not run the tool stack at all, we build the guest lists, verify them and send the invitations for you, on accounts you own. Get guest lists built without the tool stack.

Frequently asked questions