Your best customers never click — and your dashboard calls them cold
A few years ago, at a large membership retailer where I led CRM and membership analytics, I pulled up the list of customers our email tool had marked as disengaged. No opens in 90 days.
A few years ago, at a large membership retailer where I led CRM and membership analytics, I pulled up the list of customers our email tool had marked "disengaged." No opens in 90 days. The plan on the table was simple: stop mailing them, then send the stragglers a win-back discount.
Something about the list bothered me. I recognised too many names on it. So before anyone acted, I matched it against who had actually renewed and bought in the last few months.
A founder asked me the same thing more recently, in plainer words. Her dashboard says her best customers are going cold — no opens in 90 days — but the same people keep buying. Which number do I trust? Trust the buying.
Opens measure the inbox, not the customer
An open tells you someone's mail app loaded a tiny tracking image. A click tells you someone tapped a link. That is all either of them measures: behaviour inside the inbox.
Neither one measures whether a customer still wants what you sell. A person can read your subject line, decide they already know what they need, and go straight to your site. Or they can ignore every mail for a year and keep buying on a schedule they set themselves. The email did its job — it reminded them you exist — without leaving a trace in the dashboard.
The dashboard hides this by the way it labels people. It calls frequent openers "engaged" and quiet ones "disengaged." But "engaged" in that report really means "clicky." It is a measure of inbox behaviour, dressed up as a measure of loyalty.
That is the trap. Once a team believes the label, it starts making real decisions from it: who gets mailed, who gets suppressed, who gets a win-back discount. The quiet buyers get treated as if they have left, when all they have done is stop clicking.
How teams do this now
Most small teams I see run some version of the same setup. The email tool sorts subscribers into tiers by recency of opens and clicks: 30 days, 60 days, 90 days. The 90-day group gets labelled lapsed. The tool suggests suppressing them, or moving them into a win-back flow with a discount.
Nobody checks the order history first. The tiers are built from inbox data alone, because that is the data the tool has in front of it. The purchase data sits in a different system, and joining the two takes a step nobody has time for.
So a customer who bought twice in the last three months without opening a single email lands in the same bucket as a customer who has not bought in a year. Both get the same treatment: fewer emails, then a discount meant to win them back. The discount goes to someone who was never leaving.
One worked example
A small pet grooming salon had this exact problem. Its email reports showed a weak click rate, and the owner was ready to write off the quiet part of the list. Then someone cross-checked the list against bookings.
Twenty customers had each spent more than $500 over twelve months. Between them, they had opened five emails and clicked one link. They still booked monthly appointments, mostly by phone. They were among the salon's best customers, and the dashboard had them down as barely there.
Judged by clicks alone, you would say one click in a year means one interested customer out of twenty. Judged by spending, all twenty were loyal. The click count was not a smaller version of the truth. It was a different thing entirely, pointing the wrong way.
A bakery owner found the same pattern from the other side. Her email click rate sat at 3.8%, below most benchmarks, and she worried the list was dying. When she matched the list against her till, 40% of her top-spending walk-in customers were on the email list and had not clicked a single link in six months. They were not disengaged. They walked in, bought, and left. The email was a reminder on the way, not a link they needed to tap.
Research from Harvard Business School points the same way: heavy-purchasing customers often click far less than average. The people who buy the most are often the people who need the least convincing, and the least convincing is exactly what a click measures.
What changed
Opens used to be a decent proxy for attention. Then Apple Mail Privacy Protection arrived in 2021, and it started loading tracking images automatically, whether the person read the mail or not. Opens got inflated, and the number stopped meaning what it used to mean.
Since then, the metric has only got noisier. Inbox tools now summarise mail before anyone opens it, so a customer can absorb your message without ever registering an open. Clicks are still real — a tap is a tap — but they capture a smaller and smaller slice of how people respond. The customers who already trust you are the ones most likely to skip the click entirely: they know what they want, and they go and get it.
None of this makes email useless. It makes the dashboard's version of email incomplete. The channel still works. The measurement of it has narrowed to the people who behave in the one way the tool can see.
This is the same lesson as the field work on measurement I did in October: three rough measurements beat one precise-looking one. Opens, clicks, and orders each tell you something. Orders are the one that tells you whether the customer is still with you. The other two tell you how the inbox is doing.
What I would do on Monday
Build your best-customer segment from orders first, then check engagement. Take everyone who bought in the last six to twelve months, rank them by how much they spent and how often they came back, and look at that list before you look at any open rate.
Then, and only then, check how those people behave in the inbox. You will find your quiet buyers sitting there: high spend, no opens, no clicks. Leave them on the list. Keep mailing them. Do not send them a win-back discount for a problem they do not have.
And before you suppress anyone for not opening in 90 days, ask one question: when did they last buy? If the answer is recent, the dashboard is wrong about them, not the other way round. Silence in the inbox is not the same as leaving. Sometimes it is the sound of a customer who already decided, a long time ago, that they are staying.
This is the kind of work I now do with B2C teams: small, honest checks on what the numbers actually mean, before the team acts on them.
The Repeat Purchase Check — a free 15-minute self-check for B2C and DTC leaders. Five checks that show where repeat buying is leaking, and whether your personalization is doing anything about it: asksaumya-consulting.kit.com/retentioncheck
Sources: VerticalResponse, 'Your Best Customers Might Never Click Your Emails' (18 Oct 2026); Apple Mail Privacy Protection (2021); Harvard Business School research as reported by VerticalResponse (Oct 2026). The worked examples are drawn from the VerticalResponse cases cited above; the opening membership-retailer story is from the author's own practice.