# Email segmentation for ecommerce: the segments a small store needs > Email segmentation means sending each email to the group of customers it fits instead of the whole list. A small online store needs only a handful of segments: new subscribers, first-time buyers, repeat buyers, VIPs, lapsed customers at 60, 90 and 180 days, and category interest, each with its own message. Canonical: https://tellnod.com/guides/email-segmentation-for-ecommerce Last updated: 2026-10-22 Author: Tellnod team ## What email segmentation means for a store A [segment](https://tellnod.com/glossary#segment) is a group of contacts chosen by a rule, such as “ordered in the last 30 days” or “bought outerwear but never knitwear”. Segmenting means you decide who an email is for before you write it, rather than sending every message to everyone. Segmentation helps in three ways. The message is more relevant, so more people click and order. Fewer people receive email they do not want, so unsubscribes and spam complaints stay low. And you can send more often overall without any one customer feeling flooded. Gmail asks senders to email only people who want their messages and to keep spam complaints below 0.10% [S1], and relevance is the most practical way to stay there. ## Segments that matter for a small store Take Ninefold, a small apparel brand selling knitwear, outerwear and basics. These are the segments worth defining, written as rules you could say out loud: | Segment | Rule | What to send | | --- | --- | --- | | New subscribers | Subscribed in the last 30 days, no order yet | Brand story, sizing guide, a reason to place an order ([welcome series](https://tellnod.com/guides/welcome-email-series)) | | First-time buyers | Exactly one order, placed in the last 60 days | Thank-you, care tips, a product that pairs with what they bought ([template](https://tellnod.com/templates/thank-you-first-order)) | | Repeat buyers | Two or more orders | New arrivals in categories they buy, without a discount by default | | VIP | Three or more orders, or your top spenders over the last 12 months | A presale before a sale opens, a thank-you note ([template](https://tellnod.com/templates/vip-early-access)) | | Lapsed | No order in 60, 90 or 180 days | Something new since their last visit, then an offer ([template](https://tellnod.com/templates/win-back)) | | Category interest | Bought or clicked knitwear, or bought or clicked outerwear | Arrivals and restocks in that category ([template](https://tellnod.com/templates/back-in-stock)) | | Location or climate | Country, region or a cold versus warm climate | Coats when it turns cold where they live, local delivery dates before holidays | Location matters more for apparel than it first seems. A Ninefold customer in Sydney and one in Toronto need opposite products in July, and a single “autumn collection” email is wrong for one of them. ## Choosing lapsed windows: 60, 90 or 180 days How long counts as lapsed depends on how often your customers normally buy. Look at the typical gap between a customer’s orders in your own data, then set the windows around it. For a seasonal apparel brand that sees most repeat orders once or twice a year, a practical split is: - **60 days:** still warm. Show what is new; no discount needed. - **90 days:** cooling. Remind them what they liked, with a light, time-limited offer. - **180 days:** at risk. A clear win-back offer, then a last message asking whether they still want to hear from you. See the [re-engagement campaign guide](https://tellnod.com/guides/re-engagement-email-campaign). Subscribers who have not clicked or ordered in a long time are a different case from lapsed buyers: if a re-engagement email does not bring them back, removing them protects your [sender reputation](https://tellnod.com/glossary#sender-reputation). [Email list cleaning](https://tellnod.com/guides/email-list-cleaning) covers how. ## RFM in plain terms RFM stands for recency, frequency and monetary value: how recently a customer ordered, how often, and how much they have spent. The full method scores each customer from 1 to 5 on all three and end up with dozens of cells. A small store does not need that much. Three buckets on two of the dimensions give you most of the value: | | Ordered in the last 90 days | No order in 90 days or more | | --- | --- | --- | | Three or more orders | VIP: reward and preview | At-risk VIP: personal win-back | | Two orders | Loyal: new arrivals | Slipping: reminder plus offer | | One order | New customer: second-order nudge | One-time buyer: win-back | Add monetary value when you have enough orders for it to separate people, for example by moving your top spenders into the VIP row regardless of order count. ## How many segments a small list needs Fewer than you think. With a list of a few thousand contacts, four to six segments you actually use each month beat twenty you defined once. A useful test: if a segment is so small that one or two orders swing its results, merge it with a neighbor until you can tell what worked. - **Under about 1,000 contacts:** start with buyers and non-buyers, plus one lapsed segment. - **1,000 to 5,000 contacts:** add VIP, the three lapsed windows and one category split. - **Above 5,000:** add location and finer category interest, and compare results per segment. Define “engaged” by clicks and orders rather than opens. Apple’s Mail Privacy Protection loads remote content in the background when a message arrives [S2], so an open is often recorded without anyone reading the email, and an “opened in the last 90 days” segment can include people who never saw your messages. ## Building segments in Claude or ChatGPT with Tellnod With Tellnod connected to Claude or ChatGPT, you describe a segment in plain words instead of building it from filters. The assistant uses `estimate_audience` to count the people who match before anything is saved, and Tellnod removes unsubscribed and bounced contacts from the count. Requests a shop owner might type: - “How many customers have ordered twice or more but not in the last 90 days?” - “Save that as a segment called Slipping repeat buyers.” - “Email people who bought outerwear last winter but no knitwear, about the new merino sweaters.” - “Tag everyone in this CSV as VIP.” Order-based rules need your order history: import it from your store, or with the Shopify connector in Claude, ask for the customers you want and Tellnod turns them into an audience. When you are ready to send, the [confirmation card](https://tellnod.com/glossary#approval-card) shows the recipient count, who is excluded and why, and the send time, and only your Confirm schedules it. The Free plan covers up to 500 contacts; see [pricing](https://tellnod.com/pricing) for larger lists, or [connect Tellnod](https://tellnod.com/connect) to try it on your own list. ## Common segmentation mistakes - **Segments too small to learn from.** A dozen recipients cannot tell you whether a subject line worked. Merge until the numbers mean something. - **Overlapping sends.** A VIP who is also lapsed and also interested in knitwear can receive three emails in one week. Decide which segment wins and exclude the others, and check [how often to send](https://tellnod.com/guides/how-often-to-send-marketing-emails). - **Discounting your best customers.** Repeat buyers and VIPs often buy at full price. Give them access and attention before a discount. - **Rules that never update.** “Bought in spring” goes stale. Use rolling windows such as “in the last 90 days” so the segment refreshes itself. - **Forgetting consent.** A segment can only contain people who agreed to marketing email. Unsubscribes and complaints stay on your [suppression list](https://tellnod.com/glossary#suppression-list) whatever the rule says. For how segments fit into a full month of email, see [ecommerce email marketing](https://tellnod.com/guides/ecommerce-email-marketing); for turning a segmented campaign into orders, see [how to get more orders from email](https://tellnod.com/guides/get-more-orders-from-email). ## FAQ ### What are the main types of email segmentation? For a store, the useful types are lifecycle stage (subscriber, first-time buyer, repeat, VIP, lapsed), purchase behavior (recency, frequency and spend), product or category interest, engagement (recent clicks and orders) and location. ### What is a good segment size? Large enough that a single order does not swing the result. For a small store that usually means at least a few hundred recipients for a regular campaign; smaller segments are fine for personal messages such as a VIP thank-you. ### What is RFM segmentation? RFM groups customers by recency (how recently they ordered), frequency (how often) and monetary value (how much they spent). A small store can use a simple version: one, two, or three or more orders, crossed with ordered recently or not. ### Should I segment by opens? Use clicks and orders instead. Privacy features such as Apple Mail Privacy Protection record many opens automatically, so open-based segments are unreliable. ## Sources - [S1] Gmail Help: Email sender guidelines: (checked 2026-10-22) - [S2] Apple Support: Protect email privacy in Mail on Mac: (checked 2026-10-22)