How do you check if your cold email is going to spam?

We learned this the slow way this month, on a campaign we were sure was landing in spam.

A spam test can give you the wrong answer because it only measures the one mailbox it was sent to. Every mail system filters differently, and a free personal account is judged by different rules than a business mailbox.

Our tests went to a free personal mailbox because it was the one we had to hand. The list was business addresses, nearly all of it. We were reading the weather in one city and cancelling a trip to another.

We made four mistakes when we tested, and each one would have produced a bad reading on its own.

The last one was caught by a person reading the test email, not by a tool. No live campaign carried a code like that, so the test was measuring an email nobody was ever going to be sent.

You find out which mail systems your list uses by looking up the public mail records for every domain on it and grouping the results. Every business domain publishes where its mail is delivered, the lookup is free, and it takes minutes.

We ran it after the days of testing, which is the wrong way round. The largest share of the list sat with one business mail provider, a second large share with another, and a further slice behind security filters that sit in front of the mailbox. Free personal accounts were close to none of it.

Some addresses sat on domains with no mail server at all. Those are guaranteed bounces, and bounces are what turn a struggling domain into a blocked one.

Run properly, the test showed every message landing in the main inbox. We sent one message from each sending domain, each worded differently and spaced apart, into a real business mailbox on the provider that held the largest share of the list. That mailbox had never received anything from us.

It took a few minutes. The sending setup was not broken. It was failing at one kind of mailbox that the list barely contained.

It proved one provider. We could not get a reading on the second large one, so we did not guess. We sent only to the recipients we had proof for and paused the rest until we can test them.

Reaching the inbox is also not the same as getting a reply. Delivery is the entry fee. The list and the message still have to do their jobs.

You cannot trust the health score in a sending tool on its own. The summary score we were shown rated every mailbox as high, the healthiest and the most damaged alike. A score that gives your best sender and your worst sender the same mark is not measuring anything.

The one number that moved with reality was the daily share of warm-up mail landing in spam. Read that, by day, for each domain.

A change we did not need to make. We raised the warm-up volume on a set of domains, their spam share jumped the same day, and it stayed high for days after the volume came back down. A domain we left alone stayed clean throughout.

Volume alone does not explain it, because our healthiest domain was sending more than the damaged ones. The sudden change is the likeliest cause, and the damage outlasted it.

Check delivery first and leave the copy until last. If you are a small team running your own outreach, this is the hour worth spending before you rewrite a single line.

This is the same order we build in, and it is laid out on the email marketing page. If delivery checks out and replies are still thin, the list is usually next, which is covered in why cold outreach runs out of leads.