Measuring Whether Email Is Working

What the standard numbers actually measure

Three figures appear in every email report and each measures something narrower than it appears.

Delivered means accepted by the receiving server, not that it reached an inbox. A message in a junk folder counts as delivered.

Opened is inferred from a tiny image loading, which many clients block or preload. It is directional and not a fact.

Clicked is reliable and measures curiosity rather than intent. Somebody clicking a link has not decided anything.

None of them tells you whether the email produced any work, which is the only question that matters commercially.

Why open rates are worse than they look

Worth understanding rather than trusting the number.

Some mail clients load images automatically for every message, which registers as an open whether or not anybody read it.

Others block images entirely, which means a genuine reader registers as not having opened it.

Both errors exist in every list, in unknown proportions, which is why open rate is useful for comparing two of your own sends and useless as an absolute figure.

A rate of thirty percent means something relative to your own previous twenty. It does not mean thirty percent of people read it.

What to measure instead

• Replies, which are unambiguous and countable.
• Enquiries in the following week, compared with a normal week.
• Bookings or orders that follow.
• Unsubscribes, which measure irritation reliably.
• How people say they heard of you, asked directly.

The second is the most practical for a small business. If a normal week produces four enquiries and the week after an email produces eleven, that is a stronger signal than any percentage.

A worked example

A joinery business sending quarterly, with an open rate they described as disappointing at twenty-two percent.

They were considering stopping.

Counting differently: the week after each send produced roughly three times their normal enquiry volume, and two of the previous year's four largest jobs had come from somebody who mentioned the email.

Twenty-two percent of four hundred is under ninety people, and that was enough to fill a substantial part of their year.

The percentage had been making a successful channel look like a failing one, because it was being compared against a benchmark from a different kind of business.

They kept sending and stopped looking at the open rate.

Asking how people heard of you

The only measurement that connects marketing to money, and almost nobody does it consistently.

Ask every new enquiry, in the first conversation, and write the answer down.

Not as a form field, which people answer badly, but as a question in conversation, where the answer is usually specific and honest.

After thirty enquiries you have a genuine picture of what produces work, which no analytics arrangement can give you.

The answers are also frequently surprising, and they tend to reveal that the easiest thing to measure was not the thing producing the work.

The attribution problem

Worth being honest about, because it limits what any measurement can tell you.

Somebody receives your email in March, does nothing, and phones in August after seeing your van.

Which produced the job. Both, probably, and no measurement system will apportion it correctly.

This is why the honest picture comes from asking people rather than from tracking, and why a channel that shows nothing in the numbers may still be doing work.

The practical response is not to chase precision. It is to notice whether the overall flow of enquiries improves, and to accept that individual attribution is approximate.

Comparing against benchmarks

Published averages are widely quoted and mostly unhelpful.

They aggregate across industries, list sizes and list qualities that have nothing in common with yours, and a small local list behaves differently from a national retailer's.

A twenty percent open rate on a list of engaged customers who all know you is a different thing from twenty percent on fifty thousand strangers.

The only benchmark worth using is your own previous send, which controls for everything the published figures do not.

The counter-case

Where the standard metrics are genuinely useful.

Comparing two subject lines on the same list, where open rate is a fair relative measure even though it is an unreliable absolute one.

Watching for a sudden change, where a rate halving between sends means something broke rather than that the content was worse.

And diagnosing deliverability, where a collapse in opens across all recipients points at filtering rather than interest.

Used as relative and diagnostic measures, they are worth watching. Used as a scorecard for whether email works, they mislead badly.

What a small list means for the numbers

A point of proportion for anybody comparing themselves to published benchmarks.

On a list of two hundred, a difference of five percent is ten people, which is well within ordinary variation.

Which means testing subject lines on a small list produces numbers that look meaningful and are not.

The useful comparison at that scale is not between two sends but across a year, and the useful measure is whether anything happened afterwards.

Small lists do not need better measurement. They need somebody to notice whether the phone rings in the week afterwards.

The one number worth tracking over time

If you keep only a single figure, keep this one.

Enquiries per month, and where each came from, recorded in a spreadsheet with one row per enquiry.

That single record answers questions no individual channel report can: whether the business is growing, which sources produce work, and whether a change made any difference.

It also survives changing tools. Analytics packages are replaced, email providers change, and a spreadsheet of enquiries and their sources goes back as far as you kept it.

Six months of that is worth more than any dashboard, and it takes thirty seconds per enquiry.

Most small businesses have never kept it, which is why questions about what is working are usually answered with an impression rather than a number.

A proportionate arrangement

1. Note your normal weekly enquiry volume.
2. Count enquiries in the week after each send.
3. Count replies, which are the clearest signal.
4. Ask every new enquiry how they heard of you.
5. Watch unsubscribes for irritation.
6. Look at opens only for sudden changes.

That is five minutes per send, and it answers the question the percentages cannot, which is whether any of this is actually producing work.

What to send in the first place is covered in the newsletter nobody opens.

Frequently asked questions

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.