Klaviyo Segmentation: The Complete Playbook for DTC Brands

Segments versus lists: the distinction that trips people up

Almost every account we audit confuses these two, so start here. A list is a mostly static group that people opt into, such as your main newsletter list or an SMS list. A segment is dynamic: you define the conditions and Klaviyo populates it continuously, adding and removing profiles in real time as they meet or stop meeting the rules.

You never manually drop someone into a segment. You describe who belongs, and the platform keeps it current. That difference matters because it changes how you send. Blasting your whole newsletter list is how brands slowly poison their deliverability. Sending to a well-defined engaged segment is how they protect it.

The four building blocks of any segment

Every segment definition in Klaviyo is assembled from a small set of condition types, combined with AND and OR logic across definition groups:

• What someone has done (or not done): metrics and events like Placed Order, Viewed Product, Opened Email, Clicked Email, over a time window and a count.
• Properties about someone: profile properties such as a category preference, source, birthday or a custom field you set at signup.
• If someone is or is not in a list or segment: membership logic, for layering and exclusions.
• Predictive analytics: Klaviyo's modelled fields like predicted lifetime value, churn risk and expected date of next order.

The core segments every DTC brand needs

You do not need forty segments. You need a tight core that covers engagement, value and lifecycle, then a handful of tactical ones on top. These are the ones we build first in every account.

Engagement segments (30, 60 and 90 day)

These are the backbone of your campaign sending. An engaged profile is someone who has opened or clicked in a recent window. Build three: engaged in the last 30 days (your most responsive audience), 60 days, and 90 days as your everyday campaign default. You widen the window only when you deliberately want reach over response.

Buyer segments

• Bought once: customers with exactly one order. This is your second-purchase target, usually the most profitable segment to work on because a repeat buyer is worth far more than a first-timer.
• VIP: your top spenders, defined by order count (three or more) or total spend over a threshold that fits your price point. VIPs earn different messaging, early access and genuine perks.
• Never purchased: subscribers who have never ordered. Great for a harder-hitting proof-led campaign, and the pool your welcome flow is meant to convert.

At-risk and lapsed

Customers who bought before but have gone quiet past their normal cycle. This segment feeds your winback flow and, when defined against your real average time between orders rather than a blanket 90 days, it is one of the most valuable groups in the account.

Category and product affinity

People who have bought or repeatedly viewed a particular collection. These power relevant campaigns: a new-arrivals email to the people who actually buy that category converts far harder than the same email to everyone.

Engagement tiers and the deliverability connection

This is the part founders underestimate. Inbox providers like Gmail and Outlook decide where you land based on how people engage with you. Keep emailing people who never open, and you train the algorithms to route you to the Promotions tab or spam, for everyone, not just the dead weight.

Segmentation is your defence. By sending the bulk of your campaigns to an engaged segment and quarantining the unengaged, you keep your engagement rate high and your reputation intact. The mechanism has three parts:

• An engaged default: most campaigns go to the 90-day engaged segment, not the whole list.
• A sunset policy: define an unengaged segment (no open or click in 90 to 120 days, having received a minimum number of emails) and route it through a sunset and suppression process so you stop mailing people who never respond.
• Smart Sending: Klaviyo's Smart Sending setting stops you over-mailing someone inside a short window, which curbs fatigue and complaints.

Predictive analytics and RFM: the advanced layer

Once an account has roughly 180 days of order history and a few hundred customers with repeat orders, Klaviyo's predictive analytics switch on. These modelled fields let you segment on the future, not just the past:

• Predicted customer lifetime value: segment your likely high-CLV customers for VIP treatment before they have spent the money.
• Churn risk and expected date of next order: catch people drifting toward lapse and trigger a save before they are gone.
• Average time between orders: the single most useful number for timing a winback, a replenishment reminder or a cross-sell.

Klaviyo's RFM analysis (recency, frequency, monetary) groups your customers into readable tiers such as Champions, Loyal, At Risk and Needs Attention. It is a fast way to see where value sits and which groups deserve their own treatment. You do not need any of this to start, the behavioural segments above carry most of the value, but it is a genuine edge once the data is there. If you want to model the upside before committing, our revenue calculator gives a quick sense of what better targeting is worth.

How to build a segment that does not backfire

The mechanics are simple. The judgement is not. Here is the process we follow every time.

• Define the job first. A campaign audience, a flow filter and a suppression list are built differently. Decide the purpose before you touch the conditions.
• Choose the blocks. Combine behaviour, properties, membership and predictive fields with deliberate AND or OR logic. Mixing them up is the most common build error, an AND where you meant OR quietly empties a segment.
• Set the time window. Anchor behavioural rules to a period: opened email at least once in the last 90 days, placed order zero times over all time, viewed a collection in the last 30 days.
• Check the size. A segment of forty people gives noisy, unreliable results. If it is tiny, fold it into a broader group or use it as a flow filter rather than a standalone send.
• Name it clearly. You will have dozens. A good naming convention saves hours later.

Putting segments to work in campaigns and flows

Segments are only useful when they drive a send. Two main places they do the heavy lifting:

Campaigns

Send to your engaged segment, then layer included and excluded segments to sharpen the audience: a new-arrivals email to category buyers, excluding anyone who bought it this week. You can also A/B test by segment to see whether VIPs and first-timers respond to different angles.

Flow filters and conditional splits

Inside flows, segments become flow filters (only keep someone in the flow if they are, or are not, in a given segment) and conditional splits (branch VIPs down a different path). This is how a single flow serves several audiences without feeling generic. If the difference between one-off campaigns and automated flows is still fuzzy, our guide to Klaviyo flows versus campaigns lays it out, and segmentation is what makes both work harder. It also underpins the abandoned cart flow, where a purchase-based split decides who exits.

The mistakes that quietly cap revenue

• Sending everything to the master list. The fastest way to erode deliverability. Almost every campaign should target a segment.
• Never suppressing the unengaged. Hoarding dead profiles for a bigger list number is vanity that costs you inbox placement.
• One-size-fits-all campaigns. The same email to first-timers and VIPs underperforms for both.
• Ignoring predictive data you already have. If the fields are populated and you are not using CLV or churn risk, you are leaving edge on the table.
• Weak data capture upstream. Segmentation can only split on what you collect, which is why the sign-up form is where good segmentation actually begins.

None of this is complicated on its own. Doing all of it, keeping it clean and using it to sharpen every send while you run the rest of the business, is the part that eats time. It is exactly the work we take off founders' plates. For Eternal Collagen we rebuilt the segmentation and flow set and generated an extra 90,247 pounds in email revenue over four months, growing the list from around 500 to over 11,000 across six live flows. We are not claiming that is typical or guaranteed, only that this is the work, done by a senior hand.