A basket can grow in two very different ways.

A customer in a supermarket may add pasta sauce, cheese, and a drink to the product they originally intended to buy. A shopper in an electronics store may purchase only one item, but choose a higher-specification model than planned. Both transactions can produce a higher basket value, yet the commercial mechanism behind them is completely different.

That difference is captured in one of the most useful equations for understanding retail basket performance:

ATV = UPT × ASP

Average Transaction Value tells retailers how much the average transaction is worth. But ATV alone does not explain what created that value. Breaking it into Units Per Transaction (UPT) and Average Selling Price (ASP) turns a headline KPI into a diagnostic tool that store managers, merchandising teams, and retail operations teams can use to understand what customers are actually doing.

Read More :

Conversion, ATV, UPT, IPC: The 4 Retail KPIs That Decide How Many Visitors Actually Buy

ATV tells you what happened. UPT and ASP explain how it happened.

The three metrics are straightforward:

ATV — Average Transaction Value

Revenue ÷ Transactions

ATV answers:

How much does the average purchasing customer spend per transaction?

UPT — Units Per Transaction

Units Sold ÷ Transactions

UPT answers:

How many products does the average transaction contain?

ASP — Average Selling Price

Revenue ÷ Units Sold

ASP answers:

How much revenue does the retailer generate, on average, for each unit sold?

Put the formulas together:

UPT × ASP
= Units ÷ Transactions × Revenue ÷ Units
= Revenue ÷ Transactions
= ATV

The arithmetic is simple. The operational implications are much more interesting.

Suppose a store generates USD 12,000 from 300 transactions.

Its ATV is:

USD 12,000 ÷ 300 = USD 40

That tells management the average basket is worth USD 40. It does not reveal whether customers purchased many inexpensive products, a smaller number of higher-value products, or some combination of both.

The same USD 40 ATV can describe very different stores

Consider three stores that all report exactly the same ATV:

From an ATV dashboard, the stores appear identical.

Operationally, they are not.

Store A might be a sporting-goods retailer where customers typically buy one primary product and one accessory. Store B could be a grocery store where customers purchase several lower-priced items in each visit. Store C might resemble an electronics or premium beauty retailer where transactions contain fewer units but each unit carries a higher average selling value.

A manager who sees only ATV loses this information.

That is why ATV should rarely be analysed by itself.

There are two fundamental ways to build a bigger basket

Imagine a store currently has:

UPT = 2.0
ASP = USD 20

Therefore:

ATV = USD 40

Now imagine UPT increases to 2.4 while ASP remains unchanged.

2.4 × USD 20 = USD 48 ATV

ATV increases by 20%.

The improvement came entirely from customers purchasing more units.

Now consider another store:

UPT = 2.0
ASP = USD 24

Its ATV is also:

2.0 × USD 24 = USD 48

The final number is identical, but the commercial story is different.

The first store improved basket depth. The second improved average unit value.

Understanding which engine is moving matters because each requires a different response from retail management.

UPT: getting more relevant products into the basket

Increasing UPT does not mean encouraging shoppers to buy random additional products.

The strongest basket-building opportunities usually come from understanding what naturally belongs with the customer’s primary purchase.

Consider several offline retail environments.

In a supermarket, a shopper buying pasta may reasonably add pasta sauce, parmesan, vegetables, or a beverage. A retailer does not need to push unrelated products; it needs to make useful combinations visible and easy to purchase.

In a pharmacy, a customer purchasing a wound-care product may also need appropriate dressings or related first-aid items. A shopper purchasing skincare might benefit from another product that belongs to the same routine.

In an electronics store, a customer buying a laptop might genuinely need a sleeve, mouse, docking adapter, or external storage. Availability of these complementary products can materially affect the final basket.

A sporting-goods store presents similar opportunities. Running shoes can connect naturally with running socks, hydration products, or other relevant equipment.

And in a department store, basket-building opportunities can cross categories entirely—from clothing into accessories, or from homeware into related kitchen products.

The common principle is relevance.

Useful ways retailers can influence UPT include:

• placing complementary products near each other;
• making frequently purchased combinations easier to discover;
• suggesting a small number of relevant add-ons;
• creating useful bundles rather than arbitrary bundles;
• training staff to understand product use cases;
• ensuring complementary SKUs are actually available;
• analysing which product combinations repeatedly appear in transactions;
• measuring whether higher UPT also improves ATV and margin.

Digitalplace.ai supports retailers as an AI partner for retail, combining customer-facing assistance, On-Shelf Availability, and inventory intelligence to improve store execution and purchasing opportunities.

ASP: the other half of the basket equation

UPT tends to receive more attention because it is easy to visualise: put more items in the basket.

ASP is less obvious but equally important.

Suppose a beauty retailer sells two units per transaction. Last month, the average selling price was USD 18, producing an ATV of USD 36.

This month, UPT remains exactly the same, but customers shift toward higher-value skincare products and ASP increases to USD 22.

The new ATV becomes:

2.0 × USD 22 = USD 44

Nothing changed in basket size. Customers still purchased two items.

But the mix of products changed.

ASP can move because of several factors:

• customers trading up toward premium products;
• customers shifting toward entry-level products;
• price changes;
• promotions and discounting;
• assortment changes;
• pack-size changes;
• seasonal product mix;
• premium products becoming unavailable;
• different categories contributing more heavily to sales.

This makes ASP particularly useful when merchandising teams want to understand whether basket-value growth is coming from more products or different products.

Higher UPT does not always produce much higher ATV

Consider the following example.

Period 1

UPT = 2.0
ASP = USD 25

ATV:

2.0 × USD 25 = USD 50

Period 2

UPT = 2.5
ASP = USD 21

ATV:

2.5 × USD 21 = USD 52.50

UPT increased by 25%.

ATV increased by only 5%.

Why?

Because the additional units had a lower average selling value, pulling ASP down.

This could happen in many offline retail situations. A pharmacy may sell more small personal-care add-ons. An electronics retailer may attach inexpensive cables to more transactions. A department store promotion might encourage customers to add discounted accessories. A grocery retailer might increase units through a multi-buy promotion dominated by lower-priced products.

None of those outcomes is automatically good or bad.

The key is understanding what created the movement.

What if UPT rises but ATV stays flat?

This is where the equation becomes especially useful.

Suppose UPT rises from 2.0 to 2.3, yet ATV does not change.

If ATV is unchanged while customers are buying more units, ASP must have declined enough to offset the increase.

That should trigger investigation.

Possible explanations include:

• customers are adding very low-priced products;
• promotional units account for more of the basket;
• discount intensity has increased;
• product mix has shifted toward cheaper SKUs;
• higher-priced products are unavailable;
• premium products are losing share;
• bundles are adding units without adding much revenue.

A retailer that looks only at UPT might celebrate.

A retailer looking at UPT × ASP = ATV asks a much better question:

Are we building bigger baskets or merely adding more units?

One additional unit can matter at scale

Suppose an offline retailer serves 140 purchasing customers per day.

If each transaction gained one relevant additional item with an average selling price of USD 8, the theoretical incremental sales opportunity would be:

140 × USD 8 = USD 1,120 per day

Across 30 trading days:

USD 33,600

That does not mean every additional unit translates directly into incremental profit. Some purchases may replace other products, promotions may reduce margin, and customers will not respond uniformly.

But the example demonstrates why basket economics matter.

Retailers often spend significant resources bringing customers into stores. Once a customer has already entered, found a product, and decided to buy, improving the usefulness and completeness of that transaction can create value without requiring another customer acquisition.

This is especially relevant in physical retail because a large part of basket building happens through assortment, product adjacency, availability, merchandising, and staff interaction.

Product availability can determine whether the second item ever reaches the basket

A recommendation cannot increase UPT if the product is not available.

Consider an electronics store where customers who buy a particular device frequently purchase a compatible accessory. The retailer may identify the relationship perfectly, place both products nearby, and train staff to recommend the accessory.

But if that accessory is repeatedly missing from the shelf, the potential basket disappears.

The same effect can occur in beauty retail when complementary products from a skincare routine are unavailable, in sporting goods when a common shoe size is missing, or in grocery when ingredients typically purchased together are inconsistently stocked.

Basket optimisation therefore connects directly with inventory execution.

The retailer should ask not only:

“Which products should customers buy together?”

but also:

“Are those products consistently available together?”

Digitalplace.ai works with retailers as an AI partner for retail by connecting demand forecasting and Inventory Optimization with On-Shelf Availability, helping stores improve replenishment decisions while identifying shelf gaps that can interrupt otherwise valuable basket combinations.

The basket problem can begin before the customer reaches the shelf

Inventory problems do not always appear as a completely empty store.

A retailer can own the inventory and still fail to convert it into a transaction.

The product may be:

• sitting in a distribution centre;
• allocated to the wrong store;
• in a store backroom rather than on the shelf;
• misplaced in another shelf position;
• temporarily unavailable because replenishment has not happened;
• recorded incorrectly in inventory systems.

This distinction matters for UPT.

A shopper may successfully purchase the main item but abandon the complementary item because they cannot find it. The transaction still appears in the POS system, so conversion remains intact.

The loss appears somewhere else:

UPT is lower than it could have been.

That is one reason basket KPIs and inventory KPIs should not be managed as separate worlds.

Do not use UPT as a target in isolation

Every KPI becomes dangerous when people optimise the number instead of the underlying business.

Imagine a store team is heavily incentivised to increase UPT.

The easiest path might be to attach inexpensive items to as many transactions as possible.

UPT rises.

Management sees green.

But suppose ASP falls sharply and ATV barely moves. Perhaps gross margin also deteriorates because the additional products are heavily discounted.

The KPI target was achieved while the commercial outcome changed very little.

A stronger measurement structure connects several metrics:

Traffic → Conversion → Transactions → UPT × ASP → ATV → Revenue

Each stage answers a different question.

Traffic: How many shopping opportunities entered the store?

Conversion: How many visitors became buyers?

UPT: How many units did each transaction contain?

ASP: What was the average realized value per unit?

ATV: What was the resulting transaction value?

Revenue: What was the combined outcome across all transactions?

This sequence makes it easier to diagnose performance instead of simply reporting it.

Four retailers can have the same ATV and need completely different actions

Consider four physical stores, each producing an ATV of USD 60.

There is no reason these retailers should pursue the same UPT strategy.

For the sporting-goods retailer, complementary accessories may be the biggest opportunity.

For grocery, UPT may already be naturally high; improving product mix or category penetration may matter more.

For electronics, forcing more units into every transaction might be inappropriate. Ensuring the right accessories are available when relevant could matter more than maximizing unit count.

For beauty retail, the opportunity might lie in building routines across complementary products while preserving healthy ASP.

Context matters.

Look at UPT and ASP by category, store, and time period

A company-wide average can hide large differences.

Suppose a retailer operates 50 stores.

Overall UPT rises from 2.1 to 2.2. That looks like a modest improvement.

But underneath the average:

• 10 stores may have improved substantially;
• 20 may be essentially unchanged;
• 15 may have declined;
• 5 may be responsible for most of the increase because of one promotion.

The same applies to ASP.

A headline number is only the beginning of the diagnosis.

Useful comparisons include:

• store versus store;
• store versus its own historical baseline;
• category versus category;
• promoted versus non-promoted periods;
• weekdays versus weekends;
• campaign periods versus normal trading;
• premium versus entry-level assortment;
• stores with strong availability versus stores with recurring shelf gaps.

The goal is not simply to discover that ATV changed.

The goal is to understand where, when, and why.

AI can connect basket performance with the operational decisions behind it

Digitalplace.ai acts as an AI partner for retail across several operational layers that can influence basket performance. Its Inventory Optimization solution can analyse demand patterns to support forecasting, automatic replenishment, and allocation decisions so retailers can position the right inventory across stores instead of relying on the same assumptions for every location. On-Shelf Availability adds the store-execution layer by identifying shelf gaps, misplaced products, planogram issues, and other conditions that can prevent available inventory from becoming purchasable inventory. At the customer-facing level, an AI Avatar can help shoppers obtain product information, compare alternatives, and discover relevant products when staff are occupied or unavailable. These technologies should not be judged simply by whether they push UPT or ATV upward. The more useful objective is to help retailers make better inventory and customer-service decisions that improve availability, support relevant basket building, and protect the economics of each transaction rather than increasing units for their own sake.

The connection is important because basket performance does not originate from a single department.

Merchandising influences what products sit together.

Inventory teams influence what inventory reaches each store.

Store operations determine whether products reach the shelf.

Employees influence customer conversations.

Pricing and promotions influence ASP.

Customers ultimately decide what enters the basket.

ATV sits at the end of all those decisions.

A practical way to diagnose ATV

When ATV moves significantly, resist the temptation to immediately conclude that customers are simply “spending more” or “spending less.”

Work through the equation.

If ATV increases:

Check whether:

• UPT increased;
• ASP increased;
• both increased;
• promotions changed;
• product mix shifted;
• premium products gained share;
• availability improved in important complementary categories.

If ATV decreases:

Check whether:

• customers are buying fewer units;
• ASP has declined;
• discounts have increased;
• higher-value SKUs are unavailable;
• assortment mix changed;
• complementary products are missing;
• one category is disproportionately affecting the average.

If UPT rises but ATV barely changes:

Look closely at ASP.

The extra units may not be creating much incremental transaction value.

The equation every store manager should remember

ATV is an important retail KPI, but it becomes much more valuable when retailers stop treating it as a standalone number.

ATV tells you the value of the basket.

UPT tells you how many units are inside it.

ASP tells you the average realized value of those units.

And the relationship is:

ATV = UPT × ASP

If ATV falls, do not stop at saying that customers spent less.

Determine whether they bought fewer products, chose lower-value products, or both.

If ATV rises, do not stop at celebrating the increase.

Determine whether it came from more relevant units in the basket, stronger product mix, different pricing, reduced discounting, better availability, or some combination of those factors.

That is what makes the equation useful.

It turns ATV from a retrospective number on yesterday’s retail dashboard into a framework for understanding what happened inside the basket—and what the retailer should investigate next.