Why Amazon Sales Can Drop While Unit Sales Are Growing

A revenue decline usually triggers a familiar reaction: something must be wrong with the advertising. But when unit sales are rising, the first question should be what customers are buying and what those sales are worth.

More units can come from a lower-priced range, deeper promotions or the absence of a premium bestseller. It can also reflect discount-driven demand that earns very little. The numbers deserve investigation before either a budget cut or a celebration.

The useful starting point is to separate volume, price and product mix. Then connect them to availability, customer behavior and contribution profit.

1. Revenue growth and unit growth answer different questions

Units describe how much product was ordered. Revenue describes the value recorded for those orders. Within a consistently defined dataset:

Revenue = Units sold × Average selling price.

That relationship explains how volume can rise while revenue falls. It does not explain why the average price changed, whether demand improved or whether the business earned more money.

Compare equivalent periods: the same marketplace, currency, sales scope and number of days. For an event comparison, align the event days as well as the calendar dates. A two-day promotion should not be judged against an ordinary Tuesday without acknowledging the difference.

Also keep units separate from orders. One customer can buy multiple units in one order. A higher unit count does not automatically mean more customers.

2. Average selling price makes the gap visible

Hypothetical example—not a client result. A seller records the following revenue and units in a consistent sales dataset:

The arithmetic is straightforward: 100 × $400 = $40,000; 120 × $300 = $36,000. Multiplying the changes gives 1.20 × 0.75 = 0.90. The price decline more than offsets the volume gain.

This is an illustrative relationship, not a statement that every Amazon report measures net realized revenue. Seller Central’s Average Selling Price is defined as Ordered Product Sales divided by Units Ordered. Depending on the report and transaction, that figure may not reflect all discounts, refunds or subsequent adjustments. Reconcile it with transaction data before treating it as money retained by the business.[1]

3. Product mix can move the average without a price cut

A blended average combines products that may have very different price points. Suppose premium products held their individual prices, but their share of units dropped. The overall average could still fall.

Build a SKU-level bridge. For each product, compare units, recorded sales, average price and share of total sales. A useful decomposition separates changes in product volume from changes in its recorded unit price. Use the same grouping in both periods, particularly when variations or bundles have changed.

Then ask which products gained share. An increase in accessories, entry-level models or smaller pack sizes can generate positive unit growth while reducing the value of the basket. Aggregate reporting can hide that shift.

4. Missing premium inventory changes what can sell

A customer cannot purchase an unavailable variant. If an expensive bestseller goes out of stock, a lower-priced alternative may absorb some demand. Other customers may leave entirely.

Inspect available-to-buy days, not just today’s inventory balance. Match stockouts, unavailable variations, suppressed listings and loss of the Featured Offer to the period being analyzed. A product can have physical inventory without a reliably purchasable offer.

Separate products available in both periods from products whose availability changed. The first group helps you evaluate continuing performance. The second shows how much of the revenue gap may be explained by a constrained catalog. Neither calculation proves where every lost customer went.

5. Promotions change both volume and economics

Promotions can increase orders while reducing the revenue retained per unit. A displayed discount, coupon, deal fee or bundle incentive can affect economics differently across reports.

Document the actual offer, its dates and the products included. Calculate the effective customer price and the cost carried by the seller. If a free accessory is included, account for its cost and fulfillment even when the dashboard does not show it as a price reduction.

Compare promotional and nonpromotional periods separately where practical. Volume that disappears when the offer ends tells a different story from repeat purchases at sustainable prices. Check returns later; the event-day dashboard is an incomplete view of the sale’s value.

6. Competition influences the mix, but needs evidence

On a marketplace, customers compare prices, reviews, delivery promises and product features together. A lower-priced competitor might make an entry-level product more attractive, or expose weak differentiation in the premium range.

Review the actual alternatives shown for important searches and product-detail placements. Record comparable specifications and delivered prices. Avoid explaining the entire revenue decline with “cheap competitors” when your own premium products were unavailable.

Rising units can reflect stronger customer interest, bargain hunting, a shift from another channel or purchases that replace higher-margin sales. Check customer and conversion evidence where available. Unit growth is a signal to interpret, not a complete demand diagnosis.

7. What to inspect in Seller Central

Start with Business Reports for comparable ordered sales and unit trends, then examine detail-page performance at the appropriate ASIN level. Add inventory and offer-status history, promotion records and transaction reports to explain the gaps.

Not every account or report exposes identical fields. ASINs and merchant SKUs are different identifiers; parent totals and child rows can also overlap. Preserve the report definitions and avoid double-counting when joining exports. Ordered sales, shipped sales and settlement proceeds are different views, not interchangeable totals.

8. Advertising performance needs the same context

A stable advertising ROAS can coexist with falling overall revenue. Campaigns may be selling the remaining lower-priced range efficiently while the business loses its premium mix.

Match spend and attributed sales to the relevant products, marketplace and report window. Advertising attribution can credit purchases after the original interaction and may include other purchased products within the applicable scope. Do not assume campaign sales will reconcile perfectly to Business Reports for the same calendar dates.[2]

Look at total sales alongside advertising results. Where measurement permits, assess changes in non-ad-attributed sales. Simply subtracting ad-attributed sales from a different sales report does not produce a clean measurement of organic demand.

For the next layer of this analysis, see ROAS vs. TACOS: Why Strong Advertising Numbers Don’t Always Mean a Healthy Business.

9. A practical diagnostic checklist

Before changing the plan, answer these questions:

1. Are the two periods comparable in duration, event timing and data definitions?
2. What happened to units, recorded sales and average price?
3. Which SKUs explain most of the gap in revenue?
4. Did individual prices change, or did product shares change?
5. Which valuable products were unavailable or difficult to purchase?
6. What offers, costs and returns affected the realized economics?
7. Did traffic and conversion change on products available throughout?
8. How did spend and attributed sales move for those same products?
9. What happened to contribution profit and cash required to support the volume?

Assign each meaningful change to evidence, an owner and a next action. A short explanation of the three largest drivers is usually more actionable than an undifferentiated export of every metric.

10. What to fix first

Fix purchase availability and measurement gaps before using more advertising to compensate for them. Restore the products that customers cannot buy, resolve offer problems and confirm that comparisons use consistent definitions.

Next, examine mix and margin. If lower-priced products are winning sustainable business, the response may be to improve their economics or strengthen premium differentiation. If discounts are buying unprofitable volume, changing the offer can matter more than changing bids.

Only then decide whether media needs to increase, decrease or shift. Protect campaigns with supported commercial value, reduce waste and define the conditions that would justify scaling. The right action follows the diagnosis; it should not be predetermined by the direction of revenue alone.

Inovadzine’s Amazon & Marketplaces work connects marketplace performance to catalog and commercial decisions. The same questions often extend into eCommerce and the analytics and systems that make the evidence usable.

Sources and reporting notes

[1] Amazon Seller Central, Sales Dashboard definitions. Official indexed documentation defines Average Selling Price as Ordered Product Sales / Units Ordered; the full help page may require sign-in. Amazon’s official Vine help also notes that Vine orders appear in Business Reports using product prices before discounts. This is a specific example of why ordered sales must not be assumed to equal net proceeds.

[2] Amazon Ads, ad attribution methodology. Consult the actual report’s definitions and window; account type and ad product affect attribution scope.