Guide
How to test a cold email before you send it
A cold email gets one chance, in a crowded inbox, from someone who did not ask for it. You can test most of what decides its fate before it goes out: whether it arrives, whether the subject earns an open, and whether the body earns a reply. Here is how, and what only a real send can tell you.
How do I test a cold email before sending it?
Test three layers. Delivery: authenticate your sending domain and send the email to test inboxes at the main providers to check that it lands in the inbox. Copy: check that the subject is specific, the first line is about the reader, and there is one easy ask. Reaction: show it to people like your recipients, real or simulated, to hear what puts them off. Then send a small pilot batch, because only real replies measure a reply rate.
Why cold emails fail before anyone reads them
Most cold emails fail in one of three places. They never reach the inbox, because the domain is not set up or the email looks like bulk mail. They reach it and the subject line does not earn an open. Or they get opened, and the body is about the sender, with a vague or heavy ask. Each failure has its own test.
- A delivery check you can run before the first real send.
- A copy checklist for the subject, the first line and the ask.
- How to hear recipients' objections before you spend your list.
- Why a pilot batch is the only real measure of reply rate.
How it works
- 01
Set up the sending domain
Make sure the domain you send from has SPF, DKIM and DMARC records, so mailbox providers can tell the email really comes from you.
- 02
Send to test inboxes
Send the email to your own addresses at the main providers, check whether it lands in the inbox or in spam, and read it on a phone.
- 03
Run the copy checklist
A specific subject, a first line about the reader, one clear and small ask, and nothing that looks like a mass mailing.
- 04
Hear the objections
Show it to a few people like your recipients, or to simulated recipients, and ask what would make them ignore it.
- 05
Send a pilot batch
Send to a small slice of the list first, read every reply, and fix the email before it goes to the rest.
Layer one: will it arrive?
Authentication comes first. SPF, DKIM and DMARC records on your sending domain tell mailbox providers the email really comes from you, and the large providers now expect them from anyone who sends in volume. Many teams send cold email from a separate domain, so a bad campaign does not hurt the main one.
Then send the email to test inboxes you control at the main providers and see where it lands. Heavy images, lots of links, link shorteners and attachments make a first email look like bulk mail. A plain message with at most one link is safer.
Volume matters too. A new domain or mailbox that suddenly sends hundreds of emails looks like a spammer. Start small and grow gradually.
Layer two: is the copy right?
The subject line has one job: get opened by the right person. Specific beats clever. A subject that names their company, their role or a recent event reads as written for them; a subject that could go to anyone reads as a campaign.
The first line decides whether the rest gets read, and it shows in the inbox preview. Make it about the reader, not about you. A line about something their team just did earns more attention than a line about your company.
One ask, and a small one. Replying to a yes or no question is easier than booking a meeting. An email that asks for several things usually gets none of them.
Read it on a phone, out loud. If it takes more than a few seconds to see the point, it is too long.
Layer three: how will recipients react?
You cannot ask your actual recipients before you email them, but you can ask people like them. A colleague or a friendly contact in the target role can tell you what would make them delete it.
A Mimiq email test does this with simulated recipients. Paste the whole email, starting with "Subject:", and describe who receives it, such as heads of operations at logistics companies. 25 simulated recipients each see it land in their inbox among everything else that day, decide in seconds whether to open it, read it and act on it, and say why. The report counts who ignored it, skimmed it, read it and would act, and ranks the objections, such as a generic opener, an unclear ask or a tone that reads like spam.
Then ask the whole group a follow-up, such as what would make them reply, or interview the person who ignored it fastest.
What a simulated test cannot tell you
It cannot measure a reply rate or pick the winner between two close versions. Mimiq's forecast has held up on headlines. In pre-registered tests on emails, text messages and ads, it was no better than chance at ranking small wording changes, and on real web page tests it did not beat the simplest guess. So use it to find the objections, not to choose between two subject lines that differ by a word.
It knows nothing about deliverability. Simulated recipients read the email you paste; they cannot tell you whether a real inbox would file it under spam. And it cannot judge your list: an email that suits the reader you described can still go to the wrong people.
The real test: a pilot batch
Send the email to a small slice of the list first, and read every reply, including the negative ones. Replies tell you what the email made people think, in their words, and the reply rate on a slice tells you more than any test before sending.
Change one thing at a time between batches, such as the subject or the ask, so you know what helped.
Stay within the rules
Cold email rules depend on where your recipients are. In the US, CAN-SPAM asks for an honest subject line, your postal address and a working way to opt out. Rules in the EU and the UK are stricter about emailing people without consent, especially individuals rather than companies. Check what applies to you before you send.
When you test an email, use a placeholder for the recipient's name rather than pasting personal data you do not need.
What each test tells you
| Test | What it checks | What it cannot tell you |
|---|---|---|
| Domain authentication check | Whether providers can verify the sender | Whether the copy works |
| Test inboxes at the main providers | Inbox or spam, and how it looks on a phone | How real recipients react |
| Copy checklist | The subject, the first line, one clear ask | Whether this audience cares |
| A peer in the target role | Real objections from someone like the recipient | How the whole list will react |
| Simulated recipients in Mimiq | Ranked objections from people matched to the recipient, in minutes | Reply rates, deliverability, or the winner between close versions |
| Pilot batch | Real replies and a real reply rate on a slice | Nothing hidden, but it spends real recipients |
Questions
How do I check if my cold email will go to spam?
Authenticate your sending domain with SPF, DKIM and DMARC, then send the email to test inboxes at the main providers and see where it lands. Keep the first email plain, with few links and no attachments, and grow sending volume slowly.
Can I A/B test a cold email before sending?
Not in a way that predicts reply rates. Pre-registered tests found Mimiq's forecast no better than chance at ranking small wording changes in emails. Hear the objections before you send, then compare versions on real pilot batches.
What should a cold email subject line say?
Something specific to the recipient: their company, their role, or a recent event. Specific beats clever, and a subject that could go to anyone reads as a campaign.
How long should a cold email be?
Short enough to read on a phone without much scrolling: why you are writing to them, what you offer, and one small ask.
Does the Mimiq email test work for newsletters too?
Yes. It works for any email: paste it with its subject line and describe the audience. The same limits apply: it finds likely objections, not open or click rates.
What does it cost?
Your first test is free: sign in with Google and it starts right away. A free account includes 50 credits, no card. After that, a 25-person email test uses 25 credits, with a $9 Starter (120 credits that never expire) or monthly plans from $49.
Keep reading
Hear which words land before you ship them.
Your first test is free: sign in with Google and it starts right away. Last checked 2026-10-04.
Test your email free