Someone who glanced at a product once and someone who left it sitting in a cart are not the same customer, and they should not receive the same email.
Shopify customer data can do more than tell you who bought, how much they spent, or when they last ordered. The segment editor can also see what people browsed, how many times they came back to a product, and whether they abandoned a checkout.
Those are not equivalent signals. How strong the signal is should decide how specific the campaign gets. A weak signal earns a broad email. A strong one earns a message about the exact product.
Start with what the customer has already bought
Purchase history is the base layer, because it tells you what customers have already shown a willingness to buy.
Shopify gives you most of this directly in the segment editor. The number of orders and the amount spent cover the basics, first order date and last order date handle timing, and the products someone has bought matter most for campaign planning, because you can filter on a particular product, a product tag, quantity and date. That lets you build an audience like "bought something tagged running, more than twice, in the last 90 days."
Two more are worth knowing about. Shopify can group customers by how recently they bought, how often they buy and how much they spend, without you having to set those thresholds yourself. It also predicts how much a customer is likely to spend and sorts them into high, medium and low, though that one only appears once a store has passed 100 sales.
These signals create the starting groups:
- A customer who bought once and has not purchased again after the usual repeat-purchase window is a good fit for a second-purchase campaign.
- A customer who buys frequently from the same category is a good audience for a new launch in that category.
- Someone who repeatedly buys replenishable products might need a reminder when they are approaching their usual repurchase window.
Purchase history should not only label the customer, it should help answer what campaign makes sense to send next.
That also means looking beyond broad groups such as "repeat buyers" or "high-value customers." Two repeat buyers behave very differently if one consistently buys skincare and the other mainly shops home products.
What purchase history cannot tell you is what someone is interested in right now. That is where the browsing signals come in, and they are not all worth the same.
The weakest signal: they looked at a category
The entry point is a collection view. Shopify can build a segment from the collections someone has browsed although on its own, this is a weak signal.
Someone can land on a collection from an ad, a search result or a link a friend sent, look at nothing, and leave. It tells you a category was in front of them, it does not tell you if they want anything in it.
That doesn’t mean it is to waste, it still has a use. A group of customers who viewed products in the same category during the last 30 days but did not purchase are worth sending a category campaign. The audience has some interest, so the campaign does not need to start from zero.
What it does not justify is a highly specific email. If you write to that group as though they are about to buy one particular product, you will be wrong most of the time. The best way to do this is to keep it broad: the category, a few representative products, a reason to come back.
A stronger signal: they came back to the same product
Here the signal is product views, and the setting that changes everything is how many of them.
One product view is close to meaningless, three views of the same product over a week is a person thinking about it. Shopify lets you draw that line yourself, so draw it somewhere deliberate rather than treating every view as interest.
This is the signal most stores leave on the table. A customer who has returned to the same product several times is considering it even though they never added it to cart, and nothing in their order history would tell you that.
Because the signal is stronger, the email can be more specific. You can name the product, you can answer the question that is probably holding them up, which is usually sizing, fit, materials, delivery time or whether it is worth the price.
A broad category email wastes what you know about this person.
The strongest signal: they got as far as a cart
Here you are building the segment from when someone last abandoned a checkout.
This is the clearest signal in the set. They selected a product, chose a variant and started to pay. Whatever stopped them happened late, which means the gap between them and a purchase is small and usually specific.
That earns the most specific email you can send. The exact product, the exact variant, and a direct address of the thing that most often breaks a checkout: shipping cost, delivery date, payment method, or second thoughts about returns.
The mistake is treating this group like any other promotional audience. A customer who abandoned a cart yesterday does not need a category newsletter.
Match the campaign to the strength of the signal
Read as a scale rather than a list, the signals line up like this:
| What the customer did | What you filter on in Shopify | Signal strength | What the email can say |
|---|---|---|---|
| Viewed a collection once | Collections viewed | Weak | A broad category campaign, if anything |
| Viewed one product once | Products viewed | Weak | Not enough on its own |
| Returned to a collection several times | Collections viewed, more than once | Moderate | Category campaign built around what they looked at |
| Returned to the same product several times | Products viewed, more than once | Strong | That product by name, answering the likely objection |
| Abandoned a checkout | Last abandoned checkout | Strongest | That exact product and variant, plus the thing that broke the checkout |
| Bought before, now browsing a new category | Products bought, plus collections viewed | Strong | Cross-sell framed around what they already own |
The pattern is consistent: The more deliberate the behavior, the more specific the email is allowed to be, and the less you need to rely on an offer to make it work.
Combine the intent signal with what you already know
Intent tells you what someone is considering, while purchase history tells you what kind of customer they are. The strongest campaigns use both.
A frequent buyer who has recently started browsing a new category is a strong cross-sell audience, because you know they buy from you and you know what they are looking at now. A first-time buyer who has not purchased again but keeps returning to one product is a better second-purchase target than a broad group of all first-time buyers.
The offer should follow the same logic:
- Customers who regularly purchase at full price may not need an incentive even when the intent signal is strong.
- Customers who mainly buy during promotions may respond differently to the same email.
The goal is not to stack as many conditions as possible, it is to choose a combination that makes the send clearer.
When the signal is too weak to act on
Not every signal deserves a campaign, and this is the part that is easy to get wrong once the filters are available.
- A single product view is not an intent signal.
- One collection visit is not a category preference.
- A cart abandoned four months ago is a lapsed customer, not a hot lead, and emailing them as though the cart is still live reads as if nobody is paying attention.
Setting the count and the date window is where the judgment sits. Two views in three days means something, two views in six months does not. The filters will happily build either audience, which is exactly why the thresholds have to be a decision rather than a default.
If removing the intent signal would not change what you send, the campaign is really just a category email, and it should be planned as one.
AI can make these campaign opportunities easier to find
Doing this manually becomes difficult as the customer base grows.
A marketer may want to look across purchase timing, frequency, category history, product history, browsing behavior, cart activity, and other signals at the same time. The number of useful combinations grows very quickly.
AI can help by continuously looking across those signals and surfacing groups that have a clear reason to receive a campaign. AI can:
- Find frequent buyers who have started showing interest in a new category.
- Identify customers approaching their usual repeat-purchase window.
- Find customers who have repeatedly viewed the same product without buying.
The useful part is not only that AI can create the segment. It can also recommend the campaign that makes sense for that group, including the product, message, offer, and timing.
The marketer still decides whether the campaign should go out.
Shopify customer data should not only tell you who your customers are or what they bought before.
Read as a scale, it tells you how close someone is to buying, and that should decide how specific the next email gets.
Nexie is a retention marketing agent for e-commerce brands. It reads your customer data across engagement, recency, category interest and lifetime value together, surfaces the audiences worth campaigning to, and builds those campaigns for your team to review and approve, inside the guardrails and marketing policies you set. See how it works →