Do not send to the whole list. Remove obvious dead addresses, send to the most recent contacts first in small batches, and treat anyone who has not heard from you in over a year as needing to re-consent.

Why sending to all of them is the wrong move

Three things happen at once, and each makes the others worse.

A proportion of the addresses no longer exist, and a burst of bounces is a strong signal to receiving providers that you are sending to a list you have not maintained.

A proportion of the people have forgotten you, and some will mark the message as spam, which is a stronger signal still.

And your domain has no recent sending history, so there is nothing to offset either.

The combined effect can be that your ordinary correspondence starts landing in junk, which is a considerable price for one email.

The consent question first

Before anything technical, there is a question about whether you may send at all.

Express consent does not expire, so anybody who actively signed up can still be contacted, provided you can evidence it.

Implied consent does expire, and a customer or enquirer from several years ago is likely outside the relevant period.

Which means part of an old list may not be contactable at all, and sending anyway is a compliance problem as well as a deliverability one.

Working out which addresses are which is the first task, and it is frequently the reason a neglected list is smaller than it looks.

Sorting the list

The fifth is the one that causes real trouble. An old export frequently predates somebody's unsubscribe, and reimporting it puts them back on the list.

A worked example

A supplier with about nine hundred addresses accumulated since 2015 and nothing sent since early 2018.

Sorting them: roughly two hundred with a clear signup record, three hundred recent enough customers, and four hundred of unclear origin or too old.

They sent to the two hundred first, in batches of fifty over a fortnight, with a message acknowledging the gap plainly.

Bounces were under two percent, complaints negligible, and about a third opened it.

They then sent to the three hundred, with the same result.

The four hundred were not emailed. Those addresses were kept in a separate file, unused, in case any of them made contact themselves.

Had they sent to all nine hundred at once, the bounce rate alone would have been enough to damage their sending reputation for months.

Removing the obviously dead

Before sending anything, some addresses can be discarded on sight.

Anything with an obvious typo. Addresses at domains that no longer exist, which is quick to check. Role addresses at companies that have closed.

A verification service can check the rest without sending, flagging addresses that no longer accept mail.

That costs a small amount and it is considerably cheaper than the damage a high bounce rate causes.

For a list of a few hundred it is worth doing; for one of fifty, manual inspection is enough.

The message itself

Acknowledge the gap, plainly, in the first line.

Say who you are, when they signed up if you know, and why you are writing now.

Then give them something useful and an obvious way to leave. Do not open with a sales message after two years of silence.

And ask explicitly whether they want to continue hearing from you, with an easy yes. That converts a stale list into a smaller live one with fresh consent.

The people who do not respond are not necessarily gone, and treating a single non-response as a removal is over-cautious. Two ignored messages is a reasonable threshold.

Sending in batches

The technical discipline that protects your reputation.

A domain with no recent history sending nine hundred messages at once looks exactly like a compromised account.

Fifty a day, then a hundred, rising over a fortnight, lets providers see a normal pattern and lets you stop if bounces are high.

Watch the first batch closely. If bounces exceed a few percent, stop and clean the list further rather than continuing.

Most email services support scheduled batches, and doing this manually over two weeks is also perfectly workable.

The counter-case

Where the right answer is to start again.

A list whose origin is genuinely unknown, where you cannot evidence consent for any of it.

A list so old that even the recent portion is beyond any reasonable period, and where the business has changed enough that the audience is no longer relevant.

And any list assembled from sources you would not describe openly, which is a category worth being honest with yourself about.

In those cases the addresses are not an asset. Building a hundred fresh ones from current customers is faster than defending a thousand you cannot account for.

What to expect from the numbers

Setting expectations, because the results look alarming compared with an active list.

Opens will be lower than you are used to, because a proportion of the addresses are dormant even where they still exist.

Unsubscribes will be higher, sometimes several times higher, and that is the point rather than a failure. Those people were going to leave whenever you next wrote.

Bounces should be low if the cleaning was done properly, and a high figure means stopping rather than continuing.

What matters is the list you have afterwards. Six hundred stale addresses becoming two hundred live ones is a considerable improvement, even though the number went down.

Afterwards

  1. Keep the cleaned list and send regularly enough that this does not recur.
  2. Record consent for everybody who responded.
  3. Suppress the ones you did not contact so they cannot be reimported.
  4. Remove non-responders after two attempts.
  5. Set a minimum frequency so the gap never reaches a year again.

The fifth is what prevents doing all of this again in three years, and it is the one part of the exercise that is genuinely optional.

None of this is quick and all of it is cheaper than the alternative, which is either a damaged sending reputation or a list you never use again.

What frequency avoids the problem is covered in how often is too often.

Frequently asked questions

Why not just email the whole list?

Bounces from dead addresses and complaints from people who forgot you both signal a poorly maintained list, and your domain has no recent history to offset them.

What is the first question?

Whether you may send at all. Express consent does not expire; implied consent does, so part of an old list may not be contactable.

Which category causes the most trouble?

Previously unsubscribed addresses. An old export often predates somebody's unsubscribe, and reimporting it puts them back on the list.

How should I send?

In batches, fifty a day rising over a fortnight, watching bounces. A dormant domain sending nine hundred at once looks like a compromised account.

What should the message say?

Acknowledge the gap in the first line, say who you are and when they signed up, give something useful, and ask explicitly whether they want to continue.

When should I start again instead?

When you cannot evidence consent for any of it, when the business has changed too much, or when the list came from sources you would not describe openly.

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.

Old list you are afraid to email?

Sorting it by consent first usually shows it is smaller than it looks, and sending to that part safely is straightforward.

Start a Conversation