Open tracking has been substantially undermined by privacy features that load tracking images automatically, inflating open rates without anyone reading anything. Clicks, replies, and unsubscribes still measure real behaviour, and on a small list only large differences mean anything at all.

Why open rates broke

An open is recorded when a tiny invisible image loads from the sender's server. That was always approximate, since a recipient reading in a preview pane without loading images did not register.

What changed is that some mail providers now load those images automatically, on the provider's servers, regardless of whether the recipient opened anything. That registers as an open.

The result is an open rate inflated by an unknown amount that varies with what mail clients your particular list uses. A rate that rose after this change reflects the change rather than anything you did.

So opens are no longer a measure of readership. They retain limited use as a rough directional signal over long periods, and comparing your open rate against a benchmark, or against a period before the change, is not valid.

What still measures something

Clicks

Somebody deliberately tapped a link, which is a real action. The most reliable engagement measure remaining.

Worth reading as clicks per message sent rather than per open, since the denominator problem affects any rate calculated against opens.

Replies

Not reported by most platforms and the strongest signal available. A message producing three replies from a list of two hundred has done more than one producing forty opens.

Counting them manually is worth the effort on a small list, and it is the number that most closely tracks actual business.

Unsubscribes

A genuine and immediate signal. A message producing a spike is telling you something specific about that message: wrong content, wrong frequency, or a list that had forgotten who you are.

Junk complaints

The one to watch most carefully, because it affects whether future mail arrives at all. Even a small rate is a problem, and the usual cause is people who do not recognise the sender.

Bounces

Hard bounces mean the address is gone and should be removed automatically. A rising bounce rate on an older list is normal and worth acting on, since sending to dead addresses damages deliverability.

The small list problem

The statistical point that governs everything above.

On a list of two hundred, a click rate moving from four percent to six percent is a difference of four people. That is noise, not a finding, and treating it as evidence about subject lines or send times produces conclusions built on nothing.

The practical consequence is that split testing is largely unavailable to a small business. Two variants sent to a hundred people each cannot produce a statistically meaningful result about anything.

What is available instead is watching for large differences over long periods. A message type that consistently produces triple the replies of another, over a year, is telling you something real. A five percent difference on one send is not.

The number that actually matters

Enquiries and work produced, which no email platform reports.

The connection has to be made by hand: noting when a customer mentions a message, or tagging enquiries that arrive shortly after a send. Imperfect, and considerably more useful than any platform metric.

A reminder message producing four bookings from a list of a hundred and fifty is a clear result regardless of what the open rate says. That is the level at which small business email should be judged.

What to look at, and how often

MetricWhen to act
Junk complaintsImmediately, at any level
Hard bouncesRemove automatically after each send
UnsubscribesWhen a single message produces a spike
ClicksCompared across message types over months
RepliesCounted per send, watched over time
OpensRarely. Directional only.
Enquiries after a sendEvery time, by hand

Deliverability, which sits behind all of it

None of the metrics matter if messages are not arriving.

The practical checks: authentication records configured for your domain, a consistent sending address, dead addresses removed, and people who never engage moved to a low-frequency group rather than sent to indefinitely.

That last one is worth doing on any list more than a couple of years old, since a large proportion of inactive addresses drags down whether the engaged ones receive anything.

The honest summary

For a small business list, most of the reported metrics are either broken or too small to interpret.

Watch complaints and bounces because they affect delivery. Count replies because they are real. Note enquiries after a send because that is the point. Ignore the open rate, and do not run split tests on a list of three hundred.

What improves results is having something worth sending to people who remember giving you their address, which is a content and list-hygiene question rather than a measurement one, as set out in how often to email without losing people.

Frequently asked questions

Why are email open rates unreliable now?

Some mail providers load the tracking image automatically on their own servers regardless of whether the recipient opened anything, which registers as an open and inflates the rate by an unknown amount.

What email metrics still mean something?

Clicks, replies, unsubscribes, junk complaints, and bounces. All of those record a real action or a real delivery outcome.

Which metric matters most?

Junk complaints, because they affect whether future mail arrives at all. After that, replies, which are the strongest engagement signal and are not reported by most platforms.

Can I split test on a small list?

Not usefully. Two variants of a hundred people each cannot produce a meaningful result, and a difference of four people is noise rather than a finding.

What should a small business actually measure?

Enquiries and work produced after a send, noted by hand. A reminder producing four bookings from a hundred and fifty is a clear result whatever the open rate says.

What should I do about inactive subscribers?

Move people who never engage to a low-frequency group rather than sending indefinitely. A large proportion of inactive addresses drags down whether the engaged ones receive anything.

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.

Judging your email on an open rate that stopped meaning anything?

We set up the measurement that connects sends to enquiries, which is the only number worth acting on at your list size.

Start a Conversation