Before you begin
- The output is a high, medium or low estimate of future spending potential.
- Your store needs over 100 sales. Only customers who have already purchased are included.
- Sidekick can help you build a customer segment, which you review and save before using it.
What is Shopify predicting?
Yes. Shopify has a feature called predicted spend tier. It uses purchase history to estimate a customer's future spending potential and groups buyers into high, medium or low tiers.
Shopify's analytics reference describes the prediction as spending over the next year. The label does not guarantee a particular purchase amount.
That gives you a starting point for a decision such as who to invite to an early product launch. You still need to choose an offer that makes sense for those customers.
Who qualifies, and what does Shopify use?
Your store must have made more than 100 sales. The calculation includes previous purchasers, so a newsletter signup with no purchase history is outside its scope. If a customer has no assigned tier, keep that separate from low.
Shopify lists four inputs from a customer's history with your store:
- How often they purchase.
- Their average order spend compared with the store average.
- How many orders they have placed.
- How recently they last ordered.
How much should you read into a tier?
| Signal | What it tells you |
|---|---|
| Past amount spent | What the customer has already spent. This is historical activity. |
| Predicted spend tier | Shopify's estimate of future spending potential, expressed as a high, medium or low label. |
| Email opt-in | Whether the customer is subscribed to marketing emails. Check it separately before choosing an email audience. |
A prediction gives you a reason to look at a group more closely. It does not establish what those customers will buy, whether they want a discount, or how they will respond to a campaign.
I would use it to narrow a question. For a new product launch, that might be: which customers in this group have bought something relevant before? For a replenishment reminder, I would also check whether enough time has passed since their last purchase.
Ask Sidekick to build a useful segment
A customer segment is a group defined by filters, such as purchase history and email subscription status. Sidekick can translate a plain-language request into those filters, show the matching customer count and open the result for review.
In Shopify admin, open Customers > Segments, then open Sidekick. Try this request: "Create a customer segment of high predicted spend tier customers who are subscribed to marketing emails."
Open the proposed segment
Review the matching count and the filters Sidekick prepared. Check that both the spending tier and email subscription condition are included.
Inspect the customer list
Read the segment description and examine the matching customers. If you edit the filters, run the query again to refresh the results.
Save it with a clear name
Use a name such as High predicted spend, email subscribers. Saving the segment keeps the audience definition; creating and sending a campaign is a separate step.
Check what the filter actually includes
Shopify documents this filter for high predicted spend customers subscribed to email marketing. If you edit the segment manually, it belongs in the query's WHERE clause; the editor supplies the surrounding query.
AND requires both conditions. A high tier alone does not tell you whether someone is subscribed to your marketing emails. The eventual email recipient count can also be lower than the segment count, for example when an address is invalid.
predicted_spend_tier = 'HIGH'
AND email_subscription_status = 'SUBSCRIBED'Start with a question you can check
Imagine you sell coffee and want to introduce a new roast. You could start with the subscribed, high predicted spend group, then check which of those customers have bought whole-bean coffee.
Before sending anything, I would look at the size of that audience and a few purchase histories. A useful segment should make sense when you look at the people inside it.
If you later test an early-access email, decide how you will judge it before sending. Keep the offer and audience definition with your notes, and compare the response with a suitable control group if the audience is large enough. A tier label on its own does not prove that the message caused a sale.
Sources and scope of this guide
Shopify's October 30, 2026 personalization article prompted this guide. It is fresh coverage of an existing feature, not a confirmed launch date. The functionality and restrictions follow current Shopify Help documentation checked November 1, 2026.
We have not generated a segment in a merchant account, tested prediction accuracy or measured campaign results. The prompt and coffee example are exercises for readers. Shopify's documentation does not quantify the prediction's accuracy or establish a guaranteed sales outcome.
These predictions use a merchant's own customer history inside Shopify. Kraviyo does not have access to that private customer data.
