Are discounts hurting my Shopify store? The 5 signals to check

Possibly — but not in the way you're checking. Most founders judge discounts by what they cost, and discount cost is almost always a reassuringly small number. The damage, when it exists, shows up somewhere else: in basket sizes, in promo-day behavior, and in what discount-hunting customers do next. Here are the five signals that reveal it early, with real numbers from a store where the discount line looked harmless.

The trap: discount cost is the wrong metric

When we analyzed a $2M+ electronics retailer on Shopify, 38% of their ~4,700 orders carried a discount — and the total cost of all that discounting was just 0.3% of revenue.

By the metric everyone watches, discounting was a rounding error. Nobody was worried. That's exactly why nobody looked closer.

Then we segmented the orders:

• Discounted orders averaged 27% less than full-price orders
• On promo-heavy days, order volume jumped about +30% — while average basket size didn't move at all

The discounts weren't costing much. They also weren't doing much — buying volume, not bigger baskets, and pulling the store's average order value down while every dashboard said promos were "working" because orders spiked.

That's what discount damage usually looks like: invisible in the cost line, obvious in the behavior — once you ask.

The 5 signals to check (before the P&L notices)

1. Basket size: discounted vs. full-price orders. The single most revealing split. If discounted baskets are consistently and significantly smaller, your promos are attracting (or creating) a different shape of order. In the case above, the gap was 27%.

2. Promo days: volume lift vs. basket lift. A healthy promotion lifts order volume and nudges basket size (people add items to qualify or to "make the most of it"). If your promo days show volume up and basket flat — as this store's did — the discount is a coupon for orders that would likely have been small anyway.

3. Discount penetration creep. What share of orders carries any discount, and is it rising? 38% penetration at 0.3% cost means lots of small codes everywhere — which trains customers to expect one. And customers train fast: 64% of online shoppers search for a coupon or discount code before making a purchase, 85% have abandoned a cart after failing to find one, and 93% of digital coupon users say an offer needs to be at least 10% off to be worthwhile (Capital One Shopping Research, 2026). Watch the trend, not the snapshot: penetration that climbs quarter over quarter usually means codes are leaking into full-price traffic.

4. Do discount-first customers ever pay full price? Segment customers by their first order. If discount-acquired buyers rarely return — or only return with another code — you're not acquiring customers, you're renting orders. (In the same store, over half of all revenue came from one-time buyers; retention, not acquisition, was the bigger lever.)

5. Margin per order, not revenue per order. A 5% discount on a low-margin category can erase the entire contribution of the order. Revenue dashboards hide this; only a margin-aware view shows whether promo orders are worth having at all.

How to catch this early with Datajar

Every signal above is a segmentation question — the kind an analyst would answer in an afternoon, and the kind that stays unasked when there's no analyst. With Datajar, you connect your store once and ask them directly, in plain English:

• "Compare average order value of discounted vs non-discounted orders over the last 90 days."
• "On my last three promo days, did basket size change or just order count?"
• "What share of orders used a discount code each month this year — is it rising?"
• "Do customers whose first order used a discount come back and pay full price?"

You get the analysis — charted, cross-referenced, explained — in seconds instead of a spreadsheet afternoon. Run the five checks once a month and discount damage never gets six months to compound. (This is exactly how the case-study findings were surfaced →)

If the signals are bad: discount differently, not less

The fix in the case above wasn't killing promos. It was redesigning them so the same budget grows baskets instead of multiplying small orders:

• Bundle mechanics — "5% off an SSD when bought with a PC bundle" (this store's bundle buyers already spent 44% above average; 79% of them left with no add-on)
• Threshold mechanics — free shipping or a discount that unlocks above a basket-size target just beyond your current average
• Segment-targeted codes — winback offers for one-time buyers instead of sitewide codes that mostly reach people already at checkout

Discounts are a tool. The data tells you whether yours is pointed at growth or just at volume — but only if you ask.

Want to run the five checks on your store this week? Connect your store to Datajar — free to start, no credit card.

Sources

• Datajar client analysis: $2M+ Shopify electronics & PC-hardware retailer, ~4,700 orders, Jan–May 2026 (anonymized; full case study)
• Capital One Shopping Research — coupon statistics (2026): 64% search for codes before buying; 85% have abandoned a cart without one; 93% expect ≥10% off