Use the economics of the product that was purchased, then subtract the campaign’s advertising spend once. The advertised SKU’s margin belongs only to the sales of that SKU. Applying it to every attributed sale can make a campaign appear comfortably profitable when the products shoppers actually bought leave almost nothing after advertising.
For an Amazon seller, the useful calculation is purchased-product contribution minus campaign spend. Keep that result separate from both company net profit and the sales the advertising actually caused. A report can connect a purchase to an ad without proving that the customer would not have bought otherwise.
This article covers seller Sponsored Products campaigns, with hypothetical US-dollar examples. It does not apply the attribution rules of Sponsored Brands, vendors or external advertising to those campaigns. The reconciliation below is an analytical workflow, not a claim that an advertising dashboard automatically performs every step.
Why can the purchased product differ from the advertised SKU?
A shopper may click an ad for one item and buy a different size, a replacement part or another product from your catalog. In Amazon’s seller Sponsored Products reporting, promoted sales refer to the advertised SKU and halo sales to other purchased SKUs. Amazon’s advertised product report definitions describe both using a seven-day window after the ad click.
That distinction changes which costs belong to the revenue. A $40 advertised item might earn $12 before ads, while a $25 accessory earns only $2.50. A campaign that sells both has a blended contribution rate determined by what customers bought, not by the item displayed in the ad.
Keep three questions separate. Which product attracted the click? Which product earned the sale? Which advertising activity incurred the cost? The first identifies the advertised product, the second determines the product economics, and the third keeps campaign spend attached to the activity you are evaluating. Combining these into a single SKU field hides useful information.
Which reports belong in the reconciliation?
Start with the advertised product report to establish campaign scope, spend and promoted versus halo sales. Use the purchased product report to investigate products bought after clicks on ads for other products. Amazon’s sponsored ads reporting guide identifies the purchased product report as a way to understand what shoppers buy and find advertising opportunities.
Record the marketplace, currency, account type, campaign and ad-group identifiers, reporting dates, attribution window and export time. Keep the advertised SKU and ASIN alongside the purchased ASIN where supplied. An ASIN identifies the catalog product; a merchant SKU identifies your offer or inventory record. They are not interchangeable cost keys.
If one purchased ASIN maps to several merchant SKUs with different acquisition costs or fulfillment methods, investigate that mapping before assigning a precise margin. The advertising export may not contain the transaction detail needed to distinguish them. Use a documented estimate or a range when a unique match is unavailable, and label it accordingly.
Choose a single reporting grain before joining files. An advertised-product row can match several purchased-product rows. A direct join may therefore repeat the same spend on every match. Keep campaign spend in a separate total, or aggregate each source to the intended campaign and period before combining them. Check that the final spend equals the original in-scope spend.
Similarly, do not add a total-sales column to its promoted and halo components. Use either the complete total or the non-overlapping components. Preserve an unexplained difference as a reconciliation exception instead of filling it with a guessed product or deleting the row.
How does purchased product mix change the answer?
Consider one hypothetical Sponsored Products campaign advertising SKU A. The campaign spends $1,000. Its mature reporting period contains the following attributed purchases. Each unit’s non-ad cost includes cost of goods sold (COGS), applicable Amazon and fulfillment fees, and an allowance for variable return and other selling costs. Taxes collected for a tax authority and fixed company overhead are excluded from this illustration.
| Purchased product | Units | Sales | Non-ad cost per unit | Contribution before ads |
|---|---|---|---|---|
| SKU A advertised item | 50 | $2,000 | $28.00 | $600 |
| SKU B accessory | 80 | $2,000 | $22.50 | $200 |
| SKU C premium item | 10 | $600 | $36.00 | $240 |
| Total | 140 | $4,600 | Not applicable | $1,040 |
SKU A sells for $40, so its unit contribution is $40 minus $28, or $12. Fifty units contribute $600. SKU B sells for $25 and contributes $2.50 per unit, giving $200. SKU C sells for $60 and contributes $24 per unit, giving $240.
The purchased-product calculation therefore leaves $1,040 minus $1,000, or $40 after advertising, before fixed overhead. The weighted contribution rate before ads is $1,040 divided by $4,600, or 22.61%.
Now apply SKU A’s 30% contribution rate to all $4,600 of attributed revenue. The calculation produces $1,380 before ads and $380 after ads. It overstates the result by $340, even though the advertising spend and sales totals are correct. The error comes entirely from using the wrong product economics.
The opposite shortcut also fails. Looking only at the advertised item gives $600 of contribution against $1,000 of spend, suggesting a $400 loss. Including the other purchased products changes that campaign-level result to a $40 surplus. Neither headline alone explains which products supported the campaign.
Should advertising cost be reassigned to the purchased SKU?
For the campaign decision, there is no need to force every advertising dollar onto a purchased product. Sum contribution from the campaign’s attributed purchased-product mix, then deduct its spend once. This preserves the relationship between the acquisition activity and the products that followed it.
A product-level management report may still require an allocation. For example, allocating the $1,000 spend by sales share assigns SKU B $434.78: $2,000 divided by $4,600, multiplied by $1,000. That is an accounting convention. It does not establish how many clicks were needed to sell B or what would happen if B received its own ads.
Choose the allocation to answer the management question, document it, and keep it consistent. Revenue share, unit share and contribution share produce different answers. None turns aggregate reporting into a verified transaction-by-transaction acquisition cost.
Also check where advertising cost already appears in your product profit report. Deducting a campaign’s full spend again from product profit that already includes ads understates profit. Start the bridge from contribution before advertising, or explicitly reverse the advertising deduction already included. Fixed overhead should likewise be treated once and identified separately.
What if the halo products have incomplete cost data?
An unknown margin is not a zero cost. In the example, suppose you have verified the $800 combined contribution from SKUs A and B but have not resolved costs for SKU C’s $600 of sales. The remaining result depends on C’s contribution rate.
| Assumed contribution rate for C | Contribution from C | Campaign contribution after $1,000 ads |
|---|---|---|
| 10% | $60 | −$140 |
| 40% | $240 | $40 |
| 60% | $360 | $160 |
Each result equals $800 plus C’s assumed contribution, minus $1,000. The range crosses zero, so the missing cost can reverse the decision. Resolve that product first. It is more important than polishing an already accurate spend total.
Check whether cost records belong to the units and period being evaluated. A current purchase price may not represent inventory sold from an older batch. A fulfilled-by-merchant offer can have different delivery costs from an FBA offer for the same catalog product. Returns can change the expected contribution after the initial sale.
If you cannot resolve the records immediately, use a conservative and a plausible case, with the missing inputs named. A campaign that fails both cases needs a different response from one whose answer depends on a small unresolved amount. Avoid displaying several decimal places when the underlying cost is only an estimate.
How do timing differences affect the comparison?
Use a mature advertising cohort rather than judging yesterday’s clicks against yesterday’s orders. Amazon’s ad campaign attribution documentation says conversions are reported against the ad-interaction date and attribution metrics remain incomplete until the applicable window has ended. Retail order reports can therefore describe a different set of purchases even when the selected calendar dates match.
For the seller Sponsored Products scope discussed here, allow the seven-day click window to close and account for reporting updates before treating the sales split as settled. This is a reporting minimum, not a guarantee that all refunds and costs are final. Save the extraction date and recheck the cohort when later information is material.
Do not force the advertising export to equal an order-date profit report by reallocating the difference to halo sales. First explain the reporting basis, product scope, currency and exclusions. A defensible reconciliation can contain a clearly labeled residual; an invented exact match is less useful.
Where can sellerboard help with this review?
sellerboard’s PPC Dashboard documentation describes a Same SKU / All SKUs ratio and estimated profit from PPC-attributed sales. The ratio can flag campaigns where the advertised product does not represent the full purchased mix. The documentation also explains that some PPC costs use historical per-order estimates.
Start in PPC → Dashboard. Select the campaign, marketplace and a custom date range whose attribution window has closed, then open the Campaigns tab. Compare Same SKU / All SKUs with PPC sales, ad spend and estimated profit. A low ratio is a reason to investigate the purchased mix before using the advertised product’s margin.
Export the filtered campaign table to CSV to preserve the totals for your review. Use Amazon’s purchased-product detail to identify the products behind the sales, map their costs, and rebuild the contribution calculation shown above. Keep campaign spend once. This lets you distinguish an apparent profit problem from a product-mix or cost-mapping problem.
Check the summary panel’s cost assumptions as well as its profit total. Promotions, refunds and Amazon fees use previous-month per-order estimates, so compare them with the costs relevant to your purchased mix. Keep unresolved costs as a range. The documentation does not establish an exact automatic cross-SKU cost-allocation method; confirm that method before treating the estimate as a precise purchased-product calculation.
Which mistakes most often distort the result?
Using the advertised margin for every sale. Rebuild contribution from the actual purchased mix, or disclose the cost range where the mix cannot be resolved.
Treating a high halo share as automatically good or bad. Other products may contribute more or less per dollar of sales than the advertised item. Revenue share alone cannot answer the profit question.
Repeating spend in a one-to-many join. Compare the reconciled spend with the campaign source total after every merge. A purchased-product breakdown should not multiply advertising expense.
Calling attributed contribution net profit or incremental profit. Fixed overhead may still be outside the calculation, and attribution does not establish which purchases the ads caused. Name the result according to what the inputs support.
What else should sellers ask?
Does halo always mean a sale of another product from the same brand?
No. For seller Sponsored Products, other products sold by the same Seller Central merchant can qualify. Vendor Sponsored Products uses different product-eligibility rules, including a same-brand scope. Check the advertiser and ad-product type before interpreting halo as a same-brand measure.
Can I use an average margin when product detail is missing?
You can use a documented estimate, but it should reflect the purchased mix as closely as possible. Test whether a plausible margin range changes the decision. An average based only on the advertised item is especially weak when halo sales are large.
Should I pause an ad that sells few units of its advertised SKU?
Not on that fact alone. Evaluate the contribution from the full attributed purchased mix, the spend and the confidence in your cost mapping. Consider a separate test before assuming the other product would perform better when advertised directly.
What should I save before making a change?
Save the report scope and extraction time, promoted and halo totals, the purchased-product cost mapping, the spend check, unresolved amounts and the decision. That record makes a later review possible when returns or attribution updates change the estimate.
The purchased products’ economics reveal whether attributed sales covered the campaign’s spend. sellerboard’s Same SKU / All SKUs indicator helps you spot campaigns that need this cost review before deciding where to spend next.