Ecommerce advertising

QA an ecommerce promotion before and after it expires

Treat promotion expiry as a coordinated release across the ad, feed, website, and checkout. Define the exact end time and timezone, assign each system an owner, prepare replacement messaging, and verify the customer path after the cutoff. A scheduled task is not evidence that the offer actually ended correctly.

A promotion does not end in one place. The discount code can expire while ads still promise the offer, the product feed can retain a sale price after the page changes, or a countdown can restart for returning visitors.

Plan expiry as a coordinated operational event. The team needs one definition of the offer, one exact cutoff, and a way to verify each customer-facing surface afterward. This process also applies to a promotion's start, when systems can become inconsistent in the opposite order.

Write the offer as a versioned record

Document included products and variants, eligible customers or markets, discount method, exclusions, minimum spend, stacking rules, and the start and end instants. Link the approved creative and destination versions.

Use a named timezone and an explicit offset for the relevant date when scheduling systems require it. “Ends Sunday” is incomplete for a store serving several countries, and daylight-saving changes can make a remembered offset wrong.

Identify who may extend or revise the offer. A client message asking to “keep momentum” should not silently renew a discount if the business has not approved its terms and economics.

Inventory every surface carrying the promise

List active ads, scheduled ads, responsive assets, product-feed fields, landing pages, banners, cart rules, codes, automated messages, and any promotional destinations. Include copied or localized variants.

For each surface, record the owner, intended transition, scheduled time, and verification method. A static image with an embedded deadline needs different handling from a dynamic price field.

Do not assume that pausing one campaign removes every instance of the offer. An evergreen campaign may reuse the same creative, or a product ad may obtain its price from a separate feed integration.

Align feed timing with the store offer

Google's sale-price effective-date specification documents a start and end range with time and timezone. It notes that omitted timezones default to UTC. Submit explicit timing appropriate to the actual offer and verify the processed values.

Keep the regular price, sale price, and landing-page presentation consistent. Use the price-mismatch audit when the source, processed product, page, and checkout disagree.

Avoid assuming that one scheduled feed value controls a discount code or a website banner. These are separate systems unless the implementation has explicitly connected them and that behavior has been tested.

Prepare the replacement experience

Decide what a customer should see after expiry. The product may remain available at its regular price, a new offer may begin, or a dedicated promotion page may explain that the offer ended and present current options.

Keep the transition clear. Redirecting an expired promise to an unrelated product can confuse shoppers who clicked before the cutoff or saw an older shared link. A relevant destination with accurate current terms is easier to understand.

Prepare replacement ads early enough for the required review and QA. The agency approval SLA separates client approval from provider review, which helps prevent an avoidable delivery gap when the old creative must stop.

Test the boundary conditions

Use the store's authorized test process to inspect the offer before, at, and after the intended cutoff. Where a staging environment can safely simulate timing, use it to test discount logic without altering live customer terms.

Include cases that expose common ambiguity.

CaseQuestion
Cart created before expiryDoes checkout apply the intended rule afterward?
Returning visitorDoes cached content still show the old promise?
Another marketAre currency and local presentation correct?
Discount stackingDoes another code create an unintended combined offer?
Variant or bundleDo exclusions and component rules remain accurate?

The expected result should come from the business's approved offer and platform capabilities, not an assumption made by the tester.

Verify execution after the cutoff

Read the actual ad statuses, processed feed fields, visible page, and checkout rule. Record whether scheduled actions completed, failed, or remain uncertain. A calendar entry or accepted job proves only that work was scheduled.

Inspect a fresh mobile session and a representative returning session where caching matters. Confirm that the price and purchase terms are coherent across the path.

If one surface fails to transition, contain the affected exposure within the account's authority and repair that system. Communicate the specific discrepancy instead of claiming the whole promotion ended successfully because the code stopped working.

Handle extensions as new decisions

If the business extends the promotion, update the versioned offer and all relevant systems. Review the economics, stock, fulfillment capacity, and any wording that implies a genuine deadline.

Do not use an automatically resetting deadline as a substitute for a real promotion policy. The FTC's advertising guidance emphasizes truthful overall presentation; review urgency claims as part of that message.

An extension can change the interpretation of the campaign results as well. Record it in reporting so a longer promotional period is not compared casually with a shorter control period.

Review the full promotion after data matures

Reconcile net revenue, discounts, contribution, returns, and any customer adjustments related to the transition. Use the promotion economics worksheet to judge the offer rather than focusing only on order volume during the countdown.

Record failures and near misses by system. Repeated stale feed values suggest an integration issue; repeated late approval suggests a planning issue. The next promotion should inherit the repaired process, with a clear owner for each transition and evidence that the customer promise matches the purchase path.

Audit product-feed price mismatches before changing ad bids

Trace a product price from the source catalog through Merchant Center, landing-page variants, structured data, and checkout to find the actual mismatch.

Set an advertising approval SLA that clients can follow

Define review deadlines, approval scope, revision rules, and launch dependencies so client approvals do not become ambiguous campaign instructions.

Model a discount before increasing ad spend

Calculate how a promotion changes contribution per order, required sales volume, allowable acquisition cost, and the advertising evidence needed to expand spend.

Design an ecommerce bundle ad test around contribution

Test a product bundle with clear offer composition, accurate catalog data, basket-level measurement, and contribution economics instead of treating higher order value as success.

Have a correction or a question about the workflow? Contact GaaS. Read our editorial standards for sourcing and example conventions.