Open the segments tab in an account that has been running for a couple of years and you will find thirty or forty of them. Most were built for one campaign, used once, and left behind. Nobody deletes a segment, because deleting things feels risky and leaving them costs nothing visible.
Meanwhile the sends go to everyone, because picking a segment takes a decision and "all subscribers" does not.
Five segments do almost all of the useful work. Build these, use them, and treat everything else as temporary.
1. Engaged
Anyone who opened or clicked in the last 60 to 90 days.
This is the most important segment in the account, and it exists because mailbox providers judge you by how the people you send to behave. Sending your regular campaigns here rather than to everyone raises your delivery for the whole list, including the people you are not sending to.
The window is a judgement call. Sixty days is right for a store that emails several times a week. Ninety is better if you send fortnightly, because you need a longer window to catch the same behaviour.
2. Bought once, and only once
People with exactly one order.
This is usually the largest untapped group in an ecommerce list, and it is the one most stores have never spoken to as a group. They already trusted you with money. They already know whether they liked it. The distance between one order and two is shorter than the distance between zero and one, and it costs nothing to close.
What you send them is not a discount. It is the thing that goes with what they bought, or the reason to come back at the right moment.
3. Never bought, still paying attention
Subscribers with no orders who have opened something recently.
These are people stuck on something. They liked you enough to sign up and to keep opening, and something is stopping them. Usually it is a question nobody answered — sizing, delivery, whether it works for their situation.
Treating this group like the general list means selling to people who have already heard the pitch. What moves them is the answer to the specific objection, which you can find by reading your own support inbox.
4. Dormant, but not dead
No engagement in six months or more, but they have opened at some point.
This is the winback pool, and it needs to be a segment before you can decide anything about it. Most stores are either emailing this group as if nothing happened, which quietly damages the deliverability of every other send, or have already suppressed them without looking.
The useful split inside it is the one people miss: someone who opened without ever clicking is a different problem from someone who stopped opening entirely. The first group is reading and not being moved. The second is gone. The same winback email cannot serve both.
5. Best customers
Top by number of orders or total spent, whichever fits how you sell.
Worth having for two reasons. It tells you who is actually funding the business, which is often not who you think. And it is the group you should never send a hard discount to, because they were going to buy anyway and you are handing back margin for nothing.
The number of people in this segment is usually smaller and more valuable than expected, and every store owner who builds it for the first time spends a while looking at it.
How to tell a segment is worth keeping
It changes what you send. If a segment exists but every email that goes to it is the same email that went to everyone, it is not a segment, it is a filter you admired once.
That is the whole test. Not how clever the definition is. Whether the existence of the segment causes a different message to be written.
The reason segmentation gets abandoned
It is not that people do not understand it. Every store owner knows they should not send the same email to a first-time browser and a customer with nine orders.
It is that acting on it multiplies the work. Five segments means five versions, five subject lines, five reviews before send, and a calendar that used to have one entry now has five. So it gets done properly for the launch campaign, then in a busy month there is one email for everyone, and then that is just how it works now.
The work does not get smaller by being organised better. It gets smaller by not being done by hand — which is most of the reason this company exists, and the honest answer to why most accounts have forty unused segments in them.
How many segments should a store actually have?
Five permanent ones, plus whatever a specific campaign needs for the week it runs. If a temporary segment is still there a month later and nothing has been sent to it, delete it. A list of forty segments is not thoroughness, it is a place where the four that matter are hard to find.
What is the difference between a segment and a list?
A list is a group people are put into and stay in. A segment is a question Klaviyo answers continuously — anyone matching the definition is in it today and may not be tomorrow. Almost everything described here should be a segment, because the whole value is that it updates without anyone maintaining it.
