← All posts Blog · Guides

How to personalise cold email at scale with Clay

For outbound teams that want emails to feel written for one person without writing each by hand. What tools like Clay are good for, a workflow you can copy, and where personalisation stops paying off.

Short answer

To personalise cold email at scale, use a tool like Clay to pull one real signal per prospect, such as a funding round, a hiring post or the technology they use, and turn it into a single relevant line. Relevance beats flattery: the right list and offer matter more than a clever first line.

What does personalisation at scale actually mean?

One line in each email that could only have been written for that prospect, produced by a workflow instead of by hand. With Clay cold email personalisation, you pull data about each company and person, and use it to write that line automatically.

It does not mean rewriting every email. The offer and the structure stay the same. The opening earns attention.

Why does relevance come before personalisation?

Because a personalised line on the wrong list is still the wrong email. If the company cannot buy, mentioning their founder's podcast will not help. Get the list and the segment right first: an email that speaks precisely to a narrow buyer group often needs no extra personalisation at all.

See how to build a target account list.

Which signals are worth pulling?

SignalWhere it comes fromLine it enables
FundingFunding data and news"Congrats on the Series A."
HiringJob boards, careers pages"Saw you are hiring two SDRs."
Technology usedTechnology data such as BuiltWith"Noticed you run on Shopify Plus."
LinkedIn activityRecent posts by the person or company"Your post on rising CAC stuck with me."
Company factsWebsite, maps data"Saw you opened a second plant in Pune."

More on using these as openers in trigger-based cold emails.

What does a personalisation workflow look like?

  1. Start with a clean, segmented list of companies and contacts.
  2. Enrich each row with the signals that match your offer.
  3. Filter. Keep prospects with a real signal, and send the rest a strong segment-level email instead.
  4. Generate the line from the signal with a tight prompt: one sentence, plain words, no flattery.
  5. Quality check a sample by hand. Fix the prompt until the lines read like a person wrote them.
  6. Verify emails and push to your sending platform.

How do you keep quality high?

  • Only use facts that are real and public. One wrong detail destroys trust.
  • Ban flattery. "Loved your website" fools nobody.
  • Keep the line short so the whole email stays under 60 to 70 words.
  • Have a fallback for rows where the data is missing.
  • Read 50 before you send 5,000.

From our campaigns: where personalisation paid off

From our campaigns

For the decor manufacturer, the domestic campaign used one personalised line per prospect and soft asks instead of calendar links, and settled at 8 to 10 qualified meetings a month with top Indian hospitality and real estate groups.

We build lists and research with a stack that includes Apollo, Clay, BuiltWith, Bitscale, maps data and scraping, and we use funding, hiring and LinkedIn signals as openers. But the graphite campaign shows the other side: there, precision of the list did the work, and a short email naming the exact parts was personal enough.

When is personalisation a waste?

When the segment is already so narrow that the email is obviously relevant, when the data behind the line is weak, or when it slows you down more than it lifts replies. Test it: send one version with the line and one without, and compare positive replies.

About the author

Aashir Kalam is the co-founder of Emporia, a done-for-you cold email agency in Delhi NCR that has sent over 1 million cold emails and booked 1,000+ meetings for B2B companies. He writes these posts from live campaign data.

Common questions

What is Clay used for in cold email?
Clay pulls data about companies and people from many sources and helps turn it into personalised lines and segments for outbound.
How much of a cold email should be personalised?
Usually one line tied to a real signal. The offer and structure stay the same.
Does AI personalisation improve reply rates?
It can when the line is relevant and accurate. It will not rescue a poor list or a weak offer, so test with and without.
What signals work best for personalised first lines?
Funding rounds, hiring, the technology a company uses and recent LinkedIn activity.
When should I skip personalisation?
When a narrow segment already makes the email clearly relevant, or when the data is too weak to trust.
Want personalisation without the busywork? We run it for you. Book a call