Table of contents:
JetMetrics App update
Article
Post of the week
JetMetrics App weekly update
⚡️ First metric tests on the new backend
As part of our ongoing infrastructure rebuild, we’ve finished the first round of tests for metric calculations.
This week, we tested Sales on the new system — and even for large stores, it loaded up to 80× faster in local tests.
Next up: moving the rest of the metrics, starting with those used in the metric map.
We’ll share an update once this is live in production.
Check out the product’s roadmap – https://jetmetrics.io/roadmap
JetMetrics is currently in a closed beta. Book a 1:1 demo to connect Shopify store.
15 hidden patterns from metrics overlap
Most segmentation playbooks obsess over “best customers” and “win-back lists.” Cute. But there’s another side of the map: people who quietly eat your margin. They chase promo codes, turn your store into a fitting room, spam support, and leave 1-star reviews.
This piece is about them — 10 toxic segments you should spot fast and handle even faster. Simple definitions, hard criteria, and practical moves.
Let’s talk about the customers who ruin the party.
1. Discount Hunters
These customers buy only with a promo code or during sales. Full price — they skip. Over time, this pushes you into constant discounts, and your margin shrinks.
🕵️ How to spot
80%+ of their orders include a discount or use a promo code.
No full-price purchases in the last 90–180 days.
Order peaks match sale periods; almost no orders between sales.
Contribution margin for this group is below your average (after discount, shipping, and returns).
🔧 What to do
Offer value instead of pure discounts: bundles, small upgrades, early access, limited drops.
Set a higher bar to trigger promos: minimum order for discount or free shipping; one-time personal codes.
Reduce how often and how big your promo codes are; turn off auto-apply coupons.
Talk more about the product than the price: quick how-tos, comparisons, real customer use.
2. Serial Returners
These customers buy often but send a big share back. In fashion it’s especially painful — some even wear items a few times and then return them. For the store, it means lost revenue, extra logistics costs, and stock that comes back in worse condition.
🕵️ How to spot
Customer-level item return rate ≥50% (across at least two orders).
Repeated returns with reasons like “fit,” “size,” or “changed mind.”
Returns requested weeks or even months after delivery.
Atypical share of items marked “used” or “unsellable” after return.
🔧 What to do
Tighten return policies (shorter windows, stricter condition checks).
Add detailed size guides, photos, and reviews to reduce “try and return.”
Flag heavy returners and limit free returns or exchanges.
Encourage exchanges or store credit instead of cash refunds.
3. Bracketers
They order many alternatives of the same item (different sizes, colors, or models) just to keep one and drop the rest. It’s not “try two or three” — it’s “order eight to ten and keep one.” Unlike serial returners, this is about excessive alternatives and a very low kept rate right away.
🕵️ How to spot
Basket with 6–10+ variants of one product or category (for example, the same sweater in many sizes or colors).
Kept ≤20–30% within 7–14 days (or not purchased if try-before-you-buy).
Return reasons skew to fit, size, or look; returns happen immediately after delivery.
🔧 What to do
Cap variants per product (for example, max 3–4 per order) or flag baskets above it.
Improve fit UX: size guides, model specs, “fits like” hints, reviews by height and weight.
Limit free return labels for duplicate variants; add a small restocking fee.
Offer fast exchanges or store credit to avoid “order ten, keep one.”
4. Single-SKU Repeaters
They keep buying one lead product over and over (often a loss leader), don’t try other items, and rarely add anything to the basket. Great for volume, bad for margin.
🕵️ How to spot
≥80–90% of this customer’s orders include the same SKU; distinct SKUs per order = 1 (often several units of it).
Margin per order stays low or negative because the SKU is a lead or loss leader.
No category expansion after 2–3 months.
🔧 What to do
Turn the lead SKU into bundles or kits.
Exclude this SKU from coupons for repeat buyers; limit promo frequency.
Apply tiered pricing (multi-buy packs) that restores margin.
5. One-and-Done
Customers who buy once and disappear. The problem isn’t the one-time order itself — it’s when this group gets big and acquisition never pays back. The goal is to find the pattern behind them and fix the cause.
🕵️ How to spot
No second order within 60–90 days after the first purchase.
Very low engagement with post-purchase messages (opens and clicks near zero).
🔧 What to do
Cut this cohort by entry product and first-order experience (delivery speed, return, rating) to find the common thread.
Add an early second-order prompt at day 14–30 with a small product-led nudge (refill, accessory, or a simple bundle).
6. Support-Heavy
Customers who contact support far more than others. Often it’s complaints, but the key problem is that they eat up a lot of support time while bringing little revenue.
🕵️ How to spot
Tickets per order are 2–3× higher than your store average.
Ticket-to-revenue ratio is out of balance (for example, >2 tickets per $100 spent).
🔧 What to do
Flag these customers in your support system so the team knows not to over-invest time.
Keep replies short and policy-based; avoid offering extra perks or exceptions.
7. Welcome-Abusers
People who keep posing as “new customers” to claim first-order perks again and again. New email each time, same person behind it. It quietly erodes your first-order economics.
🕵️ How to spot
Multiple “first orders” linked by shared signals: same address, phone, card, device, or IP.
100% of their orders use a welcome code or referral bonus.
🔧 What to do
Make welcome codes one-time and non-stackable, bound to phone and payment method (not just email).
Set per-household limits and auto-flag matches on address, phone, card, or device.
Add light verification for first-order perks (SMS check or card match).
8. Cancel-Happy
Customers who regularly cancel after placing orders or simply don’t pick them up. Unlike returners, they never complete the purchase — so the store eats the fulfillment and logistics costs with no revenue to cover them.
🕵️ How to spot
Cancellation or no-show rate per customer is 2–3× higher than the store average.
Repeated cancellations on the same SKU or during the same promo period.
🔧 What to do
Flag high-frequency cancellers and require prepayment or a deposit before shipping.
Set stricter cut-off times for cancellations once an order is processed.
Reduce promo eligibility or perks for customers with a history of no-shows.
9. Loyalty Gamers
Customers who look loyal but only show up when loyalty points or rewards are available. They wait to spend bonuses, avoid full-price orders, and keep the program as a subsidy rather than a driver of real retention.
🕵️ How to spot
80–100% of orders are paid partly or fully with loyalty points.
Order activity spikes right after bonuses are issued; near zero in between.
Contribution margin per order is below average once points and rewards are applied.
🔧 What to do
Adjust the program so rewards unlock with higher spend or multi-category baskets.
Cap how much of an order can be paid with points.
Offer non-discount perks (early access, exclusive drops) to shift focus away from pure price.
10. 1★ Amplifiers
They don’t spend much, but make a lot of noise: constant 1-star reviews, complaints, or negative posts on social media. The damage goes beyond their own orders — they drag down product ratings and scare off new buyers.
🕵️ How to spot
Low total spend but a high volume of negative reviews or tickets.
Around 90% of their reviews are 1–2★.
🔧 What to do
Moderate negative reviews from these customers first; try to resolve privately before publishing.
If reviews are repetitive or abusive, push them down (end of list) or remove them when possible.
Add clear seller responses under their reviews to reduce impact and show context.
Final thoughts
Not every customer is a good customer. Some groups look active but quietly cost more than they bring. That’s why segmentation isn’t just about VIPs — it’s also about naming the segments that drain margin and setting clear boundaries for them.
LinkedIn post of the week
Happy analyzing 🫶
See you next week!
Dmitry from JetMetrics