Tiered Discounts on Shopify: Thresholds That Raise AOV
A threshold set at a round number rewards customers who were already spending that much. Set it from your own distribution and it changes behaviour instead.

On this page
Key takeaways
- Thresholds should come from your actual order value distribution, not a round number, or you discount behaviour you already had.
- The progress indicator is the mechanic — a shopper who cannot see how close they are has no reason to add anything.
- Too many tiers reduces response; two or three rungs outperform five in most catalogues.
- Measure revenue per visitor, not average order value, because AOV rises whenever a threshold exists regardless of whether you earned anything.
A spend threshold is one of the better mechanics available to a Shopify merchant, because it raises order value without reducing unit margin. The customer who was going to spend below it either adds something or does not; the customer above it pays what they were going to pay.
It also fails quietly and often, for two reasons that are both fixable: the threshold is set at a number that means nothing, and the shopper cannot see how close they are.
Why the mechanic is attractive
A percentage discount costs you margin on every order. A threshold costs you nothing from shoppers who do not reach it, and costs you a defined amount from those who do — while changing what they buy rather than what they pay per unit.
Set the threshold from your own data
The most common failure is a round number. ₹2,000, ₹5,000, £50. Round numbers are chosen because they look tidy and they bear no relationship to how your customers actually spend.
The number you want comes from your order value distribution. Export your orders, and look at where they cluster.
Slightly above the cluster is where the threshold belongs. If most of your orders land between ₹1,400 and ₹1,800, a threshold at ₹2,000 puts a meaningful share of shoppers within one item of qualifying. That is a reachable step, and reachable steps get taken.
Far above the cluster and the offer is invisible to most of your traffic. It only rewards customers already spending that much, which is a discount on existing behaviour.
Below the cluster and almost everyone qualifies without changing anything, which is simply a discount with extra steps.
The right threshold usually sits somewhere between ten and thirty percent above your median order value, but the range matters less than the principle: derive it, do not choose it.
One refinement worth making if your catalogue spans very different price points. A single threshold that works for your mid-range products may be trivially cleared by anyone buying a premium item and unreachable for anyone buying accessories. Where that gap is large, either set the threshold against the segment that represents most of your orders and accept it does little for the others, or use category-scoped thresholds — which is more work and more explaining, so only worth it when the spread genuinely warrants it.
A worked example of setting the number
The principle above is straightforward to agree with and easy to skip in practice, so it helps to walk through the actual steps on a hypothetical order file.
Export the last three months of orders, excluding any period distorted by a major sale, since a threshold planned around discounted-period baskets will sit at the wrong level for the rest of the year. Sort the order values and look at the distribution rather than the average alone, because an average can be pulled upward by a handful of large orders while most customers cluster well below it. A median is usually the more honest single number to start from.
Say the median order value comes out around a certain figure and the bulk of orders, the middle band that excludes the smallest and largest tails, cluster in a range running from somewhat below that median to somewhat above it. The threshold candidate sits above that cluster's upper end, not above the median alone, because the goal is a number that a meaningful share of shoppers can reach by adding one more item, not a number only your highest spenders already clear.
Cross-check the candidate against your typical single-item price. If the gap between the median order and the proposed threshold is smaller than your cheapest regularly sold item, almost nobody will need to add anything to qualify, which defeats the mechanic entirely. If the gap is larger than your most expensive typical item, reaching it requires adding two or three things, which is a harder ask and will show up as a lower response rate.
Once live, the number is not permanent. Revisit it every few months against fresh order data, because product mix, pricing changes and seasonal shifts in what people buy all move the distribution the threshold was built from. A threshold that was well placed in spring can quietly become too easy or too hard to reach by autumn if nobody checks it again.
The progress indicator is the mechanic
This is the part merchants skip, and skipping it is why thresholds underperform.
A threshold communicated as a site banner — "Free shipping over ₹2,000" — is a fact the shopper reads once and forgets. At the moment they are deciding whether to add another item, it is not in front of them.
What works is distance, shown in the cart, in money, updated as the basket changes:
"You are ₹340 away from free shipping."
That sentence does the work. It is specific to this shopper, it names a small number, and it appears exactly where the decision is being made. The generic banner does not.
Native Shopify discounts do not render this — they apply silently until the cart shows the result. Building it is either a theme change or the main reason to use an app in this category, and it is worth being clear that the progress indicator, rather than the discount logic, is what you would be paying an app for.
Where the threshold needs to be visible
- Cart drawer or cart page, showing distance in money
- Updated live as items are added or removed
- Stated again at checkout, confirming it applied
- On product pages for qualifying products, at least as a line
- In the announcement bar, as the general fact
- Never only in the announcement bar
Fewer rungs, not more
The instinct is to build a ladder — five percent at one level, ten at the next, fifteen above that. The instinct is usually wrong.
Every additional tier turns a simple decision into a table to study, and shoppers do not study tables. Response falls as complexity rises.
Two or three rungs is the practical range. Often one threshold is enough. A second rung, set at a genuinely aspirational level, catches the customers who would stretch and does no harm.
Beyond three, you are managing complexity for no gain and creating more edge cases where offers interact badly. Each additional rung also multiplies the testing burden, since every tier boundary is a place where the reward can apply incorrectly and nobody will notice until a customer does.
Choose the reward carefully
What sits on the other side of the threshold matters as much as where it sits.
Reward types compared
| Reward | Cost to you | How it reads |
|---|---|---|
| Free shipping | Fixed, known | High perceived value, especially on smaller baskets |
| Free gift | Cost of the item | Generous relative to cost; needs inventory planning |
| Percentage off | Scales with order value | Costs most where least needed |
| Fixed amount off | Fixed | Simple, easy to communicate |
| Loyalty points or credit | Deferred | Cheap now, encourages return |
Free shipping is usually the strongest option. It is a fixed and known cost, it reads as valuable to shoppers, and it removes a friction that already exists rather than adding a benefit that did not.
A percentage is the weakest, because it costs you most on the largest orders — exactly the customers who needed the least persuasion.
Where it quietly fails
Four failure modes worth checking before launch.
Nobody can see it. Covered above, and by a considerable distance the most common of the four.
The threshold is calculated on the wrong number. Before or after discounts, including or excluding tax and shipping. Decide, be consistent, and state which — a customer who thought they qualified and did not will contact support, and they will be right to.
It interacts badly with other offers. A shopper qualifying for both a threshold reward and another discount hits Shopify's combination rules. Test the pairing rather than assuming.
The basket drops below after qualifying. The reward should be removed when the cart no longer qualifies. If it is not, you are shipping free products or free delivery on unqualified orders, and the pattern gets shared once someone notices.
Help the shopper close the gap
Showing the distance creates the intent. What happens next determines whether it converts, and most stores leave the shopper to work it out alone.
A customer told they are ₹340 away now has a small problem: what costs roughly ₹340 and is worth having? If solving that means going back to a collection page and browsing, a good share will not bother.
Suggest something in range. A small set of products priced near the gap, shown in the cart. This is the single highest-return addition to a threshold mechanic and it is frequently absent.
Price the suggestions slightly above the gap, not below. Something at ₹380 closes it; something at ₹300 leaves them ₹40 short and mildly annoyed.
Favour genuinely useful additions. Consumables, accessories, refills — things a customer would plausibly want rather than whatever is nearest the price point. Suggestions that look arbitrary undermine the offer.
Keep it to three or four options. This is a nudge, not a browsing experience.
There is a related decision about what counts toward the threshold. Excluding sale items is defensible and needs saying plainly on the offer, because a shopper who assumed otherwise discovers it at the least convenient moment. Whatever you decide, state it where the offer is stated rather than in a policy page nobody reads.
Round the gap in the customer's favour
If the shopper is ₹7 away, apply the reward. The goodwill from that is worth considerably more than ₹7, and the alternative is a customer who adds a cheap item purely to clear a trivial gap and feels manipulated doing it.
Measure the right number
Average order value will rise. It rises whenever a threshold exists, because qualifying orders are by definition above it. This tells you nothing.
The question is whether revenue per visitor rose. That captures both what customers spent and whether the offer cost you any conversions.
Compare against a period before the threshold, or better, a group of visitors not shown it. And watch return rates, because items added to reach a threshold are chosen with less deliberation than items sought out, and they come back more often.
Watch what happens when you remove it
A threshold offer that runs permanently becomes the expected price. Customers learn the number and shop to it, which is fine while it runs and produces a visible drop when it stops. If you intend it as a permanent mechanic, that is a decision worth making deliberately rather than discovering when you try to end it.
For the wider offer strategy and margin arithmetic, see promotions and offers. Implementation and combination rules are covered in the same archive, and BFCM-specific structures in the BFCM 2026 series.
Frequently asked questions
- Where should I set a spend threshold?
- Slightly above the point where most of your orders currently land, so a meaningful share of shoppers are close enough for one more item to be a reasonable step.
- How many discount tiers should I have?
- Two or three. Response falls as tiers multiply, because the offer stops being a simple decision and becomes a table to study.
- Do tiered discounts actually raise average order value?
- They raise it mechanically, since orders that qualify are by definition above the threshold. Whether they raise revenue is a different question and needs a comparison against a period or group without the offer.
- Why is nobody hitting my threshold?
- Usually because they cannot see it. A threshold communicated only as a banner has no effect at the moment of decision.
- Should the reward be a percentage or free shipping?
- Free shipping is usually cheaper for you and reads as more valuable to the shopper, particularly on lower-value baskets.



Comments