Free Listings, Promotions and Automated Discounts in Merchant Center

by Francis Rozange | Apr 4, 2026 | Google Ads

Free listings, promotions, and automated discounts: the part of Merchant Center that does not bill you per click

Most merchants treat Google Merchant Center as the silent plumbing behind Shopping ads. They feed their catalog, fix the disapprovals, and forget that three of the most consequential growth levers in the platform have nothing to do with paid auctions: free product listings on Surfaces Across Google, the Promotions feature that paints a colored badge under your tile, and Automated Discounts, the machine-learning layer that adjusts the displayed price in real time based on conversion likelihood.

None of these features bill you per click. All of them depend on the same plumbing: a clean feed, a landing page that matches your data, and a strict reading of Google’s commercial policies. The merchants who win at Merchant Center are not the ones bidding hardest. They are the ones whose feed quality lets free listings, promotions, and automated discounts compound on top of paid traffic.

This guide covers what each of those three features actually does, the eligibility rules nobody reads until they trigger a disapproval, the relationship between the sale_price attribute and the Promotions card, the mechanics of automated discounts, the recurring rejection patterns, and a measurement playbook that distinguishes promotion lift from baseline drift.

Free listings: Surfaces Across Google explained

The feature now called free product listings was rolled out in late April 2020, when Google announced that the main Shopping tab would become predominantly free for merchants. Search Engine Land covered the rollout at the time as the most significant Shopping product change since the platform pivoted to a paid model in 2012. The shift was framed as a response to the COVID-19 e-commerce surge, but it was permanent: free listings stayed, and they have since expanded to other surfaces.

Today, the program Google calls Surfaces Across Google distributes eligible products across:

The Shopping tab in Google Search.
The main Search results page, in product result modules.
Google Images, where product tiles appear annotated with price and merchant.
Google Lens results, when a user photographs an item.
YouTube, in shopping ads adjacent to creator content and in the Shopping shelf.
Google Maps, for merchants with a verified Business Profile and a local inventory feed.
Conversational AI surfaces in Gemini, when shopping intent is detected.

None of those placements charges you. The visibility is paid for by the same feed quality and policy compliance that earns you good Quality Score on the paid side. Search Engine Journal has documented several cases since 2021 where merchants who had been spending exclusively on Shopping ads turned on free listings and saw 15 to 35 percent additional clicks per month with no incremental cost, drawn primarily from the Shopping tab and from Google Images.

Eligibility: what Google actually checks

Free listings are not automatic. Google enforces a layered set of conditions before a product becomes eligible:

An active product feed in Merchant Center, refreshed at the cadence required by your inventory volatility (daily for most merchants, hourly for high-velocity SKUs).
An account in good standing, meaning no current policy suspension and no warning state on the account dashboard.
No item-level disapproval blocking the SKU. Free listings inherit the same item-level checks as paid Shopping: GTIN validity, image policy, landing page reachability, price match between feed and site.
Compliance with the Shopping ads policies, which extend to free listings. Categories restricted in paid Shopping (alcohol with restrictions, financial services with disclosures, etc.) are also restricted in free listings.
A verified and claimed website. The verification step links your domain to your Merchant Center account and prevents one merchant from claiming another’s catalog.
For Google Maps placement, a linked Business Profile and a local inventory feed indicating in-store availability.

Google publishes the full eligibility matrix in the Merchant Center Help center under Surfaces Across Google. The rule that catches the most merchants by surprise is the policy parity between free listings and paid Shopping: a single disapproval on a SKU removes it from both at the same time. Merchants who think free listings will rescue a catalog with chronic feed quality issues are wrong. The opposite is true: if you have not earned eligibility on the paid side, you will not be visible on the free side either.

Free listings versus paid Shopping: not a zero-sum game

The most common objection from agencies in 2020 and 2021 was that free listings would cannibalize paid Shopping clicks. Tinuiti, the largest Google Premier Partner that publishes regular industry benchmarks, ran the numbers across hundreds of accounts in 2021 and 2022 and found the opposite: free listings added incremental traffic, mostly from queries and surfaces where paid Shopping was not bidding aggressively. Search Engine Land summarized the Tinuiti findings: paid Shopping CTR did decline marginally on the Shopping tab where free listings now appear immediately below the paid carousel, but the decline was offset many times over by the new free traffic.

The mental model that works:

Paid Shopping captures high-intent queries with large auction prices and wins the top of the page.
Free listings catch the same queries lower on the page, plus the long tail of queries where you do not bid, plus the discovery surfaces (Images, Lens, YouTube) that paid Shopping rarely covers.
The two channels share a feed and a Quality Score signal, so investing in feed quality benefits both at once.

Treating free listings as paid traffic stealers is a category error. They are a parallel distribution layer that uses the same plumbing.

The Promotions feature: a badge that doubles as a CTR driver

Promotions in Merchant Center are not a discount engine. They are a metadata layer that tells Google to display a colored badge (“Special offer”, “Sale”, “10% off”, “Free shipping”) under your product tile in Shopping results, in the Search shopping module, and in some surfaces of YouTube Shopping. The actual discount has to be honored on your landing page and at checkout. The promotion attribute simply earns you the badge.

The badge has measurable impact. Tinuiti’s promotion analyses, replicated by Smart Marketer and Store Growers in their respective audit decks, consistently show CTR lift in the range of 8 to 25 percent on tiles carrying a promotion badge versus baseline tiles for the same SKU. The lift is largest on price-sensitive categories (apparel, accessories, consumer electronics) and smallest on commodity categories where price is already the only differentiator.

Promotion types Google supports

The Promotions feature supports four mechanical types:

Percent off, expressed as a percentage applied to a defined product set or to the entire cart. Example: 15 percent off all wool sweaters.
Amount off, expressed in your feed currency. Example: 10 dollars off any order over 50 dollars, or 25 euros off the listed product.
Free gift, where a complementary item is added at checkout when the qualifying cart is built. Example: a free travel pouch with the purchase of any backpack.
Free shipping, either standard or expedited, with or without a minimum order threshold. Example: free 2-day shipping on orders over 75 dollars.

Each type has its own attribute schema in the promotion feed. The free shipping type has its own surfacing on the tile (a separate “Free delivery” annotation), distinct from the percent-off and amount-off badges. Free shipping promotions also benefit from a higher visibility weight on mobile Shopping results, where shipping cost is a leading abandonment driver.

Where promotion badges appear

The badge surfaces in the following placements, ordered roughly by traffic share for most merchants:

The Shopping tab in Google Search, both paid and free results.
The Shopping module on the main Search results page (the row of product tiles below or beside the blue links).
Google Images, where product results carry the badge below the merchant name.
YouTube Shopping, in the product shelf that creators can attach to videos.
The Gemini Shopping experience, when promotional offers match the conversation.

The badge does not appear on the merchant’s own website. It is a Google-side annotation only. If you want the same messaging on your site, you have to render it yourself in your theme. This is a recurring source of confusion for merchants who expect the promotion to flow through to their storefront.

Setting up a promotion feed: dates, offer codes, and product targeting

Small merchants can create individual promotions in the Merchant Center UI under Marketing > Promotions. Larger catalogs need a promotion feed, which is a separate file from the product feed and uses its own schema documented in the Merchant Center Help center.

The minimum required attributes for a promotion feed entry:

promotion_id: a unique identifier you control. Used to map products to promotions and to update the promotion later. Reusing a promotion_id replaces the existing promotion; using a new one creates a new promotion.
product_applicability: either “all_products” or “specific_products”. The latter requires you to declare which SKUs the promotion applies to via item_id, item_group_id, product_type, or brand filters.
offer_type: “no_code” if the promotion auto-applies, or “generic_code” if a single shared code is required at checkout.
generic_redemption_code: the actual code, if offer_type is “generic_code”. The code must work on your site for any user.
long_title: the full text of the promotion, as it would be read aloud. Example: “Save 15 percent on all winter coats with code COLD15”.
promotion_effective_dates: an ISO 8601 interval, with start and end timestamps including timezone offset. Format: 2026-04-15T00:00:00-05:00/2026-04-30T23:59:59-05:00.
promotion_display_dates: the window during which the badge can be shown on Google surfaces. Usually identical to effective_dates, but can be set earlier to surface the offer before it activates (for pre-promotion teasers, where Google supports them).
redemption_channel: “online”, “in_store”, or both. Most merchants use online only.
percent_off or money_off_amount: the actual discount. Money_off_amount carries a currency.

The dates are the field that catches the most merchants. Google enforces a strict reading of the timezone offset. A promotion declared in UTC that should run on Pacific time will end seven or eight hours early. The Help center recommends always declaring the offset explicitly and aligning it with your e-commerce platform’s timezone setting.

Offer codes deserve their own attention. The generic_redemption_code is shown on the badge as part of the long_title and copied to the clipboard automatically on most surfaces. If the code does not work on your site, the promotion is rejected on the next review pass. Codes that require a minimum cart amount have to declare that minimum via minimum_purchase_amount; if you forget, customers triggered by the badge will fail to redeem and your promotion will be flagged for inconsistency.

Automated Discounts: the ML layer that adjusts the displayed price

Automated Discounts is the most recent of the three features and the one merchants understand the worst. It is opt-in by feature, off by default, and it does not change the price the customer pays at checkout. It changes the price Google displays on the Shopping tile, the Search module, and the Surfaces Across Google placements, based on Google’s machine-learning estimate of conversion likelihood for that specific user, query, and context.

Google’s stated goal with Automated Discounts is to maximize gross profit, not to maximize click volume. The merchant provides:

The cost of goods sold attribute (cost_of_goods_sold) at the SKU level.
An optional minimum_displayed_price floor, below which the algorithm cannot go.
An opt-in toggle in Merchant Center under Growth > Automated Discounts.

Google’s ML then evaluates each impression, predicts how much discount is needed to convert that user, and renders the displayed price accordingly. A user with high conversion likelihood (return customer, high-intent query, mobile session in a high-converting hour) might see the full listed price. A user with lower predicted likelihood, or one with high price sensitivity signals, might see a 5, 10, or 20 percent reduced price.

The reduced price is then applied at checkout via a coupon or cart adjustment that the platform handles. The customer never sees a different price than what was advertised on the tile. Search Engine Land’s coverage of the Automated Discounts launch and subsequent updates noted that the feature is most effective on accounts with at least several hundred conversions per month, since the underlying ML needs sufficient signal to differentiate users.

Tinuiti’s testing of Automated Discounts across a sample of mid-market merchants in 2024 reported displayed-price reductions averaging 4 to 12 percent, with conversion lift of 6 to 18 percent and gross profit lift of 3 to 9 percent versus a holdout group. The variance across categories is large; apparel and home goods saw the highest lift, while categories with thin margins and fixed price points saw little movement.

The relationship to sale_price: who is the source of truth

Automated Discounts does not override the sale_price attribute. It complements it.

sale_price is the merchant-declared discounted price. It applies to all users, all the time, for the duration declared in sale_price_effective_date. It is the source of truth for the customer-facing offer.
The Promotions feature attaches a badge to the tile referencing a defined offer (percent off, amount off, free shipping, free gift). It does not change the price displayed in the price field.
Automated Discounts adjusts the displayed price below the listed price (or below sale_price if it is set), on a per-impression basis, based on ML signals. The merchant declares COGS and a minimum floor; Google does the math.

The cleanest mental model: sale_price is always your floor for the customer-facing offer. Promotions add a badge. Automated Discounts adds a per-impression adjustment that can go below sale_price down to your declared minimum_displayed_price. If you do not declare a minimum, Google uses COGS plus a small margin.

If you set a sale_price and also enable Automated Discounts on the same SKU, the displayed price will be the lower of the two for any given impression. The customer pays the displayed price. Your reporting in Merchant Center shows both the sale_price and the algorithmic adjustment as separate columns, so you can isolate the contribution of each.

Why promotions get rejected: the recurring patterns

Promotion rejection rates are higher than product feed rejection rates because Google reviews promotions more strictly. The reviewer is checking three things: that the offer is real, that the offer is honored on the landing page, and that the offer language is not misleading. The recurring rejection patterns:

Vague or unverifiable terms

Promotions declared with language like “Save big” or “Limited time offer” without a quantified discount are rejected on first review. Google requires the long_title to specify the discount amount or percentage and, if applicable, the qualifying conditions. “Save 15 percent on all backpacks” is approved. “Big savings on backpacks” is not.

Missing or incorrect dates

Promotions without promotion_effective_dates are rejected. Promotions with dates that have already passed at the time of submission are rejected. Promotions with timezone offsets that contradict the merchant’s declared business location are flagged for review and often rejected for inconsistency.

Landing page mismatch

This is the most common rejection. Google’s reviewer fetches the landing page and verifies that the promotion is actually claimable. If the badge says “15 percent off with code COLD15” and the landing page does not display the code, does not have a coupon field at checkout, or rejects the code at cart, the promotion is disapproved. Smart Marketer’s audit playbook for promotion compliance puts landing page parity as the single largest source of rejection across the merchants they audit.

Inconsistent discount math

Promotions where percent_off does not match the actual discount applied at checkout are rejected. Example: a promotion declares 20 percent off, but the cart applies 15 percent. The reviewer catches this and disapproves.

Stacking conflicts with sale_price

Promotions that, when applied to a SKU already discounted via sale_price, would result in a final price below the manufacturer’s minimum advertised price are rejected for some categories (consumer electronics in particular). This is rare but worth knowing if you sell branded electronics.

Restricted categories

Some categories cannot run promotions at all in some markets: alcohol in markets with promotional restrictions, prescription healthcare, financial services with regulated disclosure requirements. The Help center maintains the list, and Google updates it as local regulations change.

Free shipping declared but not actually free

A free shipping promotion has to mean free shipping. If your checkout still charges a shipping fee under any condition (taxes, handling, surcharge for some zip codes), and your promotion does not declare those exceptions, the promotion is disapproved. The clean fix is to declare the minimum_purchase_amount and any geographic exclusions explicitly.

Measurement playbook: isolating promotion lift

The hardest part of promotions is not setup. It is measurement. A promotion launched on a high-traffic week will look successful regardless of whether the badge actually moved any customers. A promotion launched in a slow week will look weak even if it added meaningful incremental conversions. Without a control, you cannot tell.

The cleanest measurement design, used by Tinuiti and Store Growers in their case studies:

Step 1: define the SKU set

Pick the products you intend to promote. Split them into a treatment group (will carry the promotion) and a control group (will not). The split should be random within product categories of similar conversion rate. Aim for at least 30 SKUs per group for stable readings, more for low-volume catalogs.

Step 2: lock the price and the inventory

The treatment and control groups should have the same effective price (no sale_price difference) and similar inventory levels for the duration of the test. Otherwise the test confounds promotion lift with price lift or stockout differences.

Step 3: run the test for a full purchase cycle

Most categories need 14 to 21 days to surface meaningful conversion data. Ending a test at day 7 because “the lift looks good” is a recipe for false positives, since promotion days early in the test tend to over-index due to repeat visitors.

Step 4: compare conversion rate, AOV, and gross profit

The three metrics that matter:

Conversion rate. Treatment minus control gives you the promotion’s CTR-to-purchase lift.
Average order value. Promotions sometimes cannibalize AOV by encouraging single-item purchases at the discount; sometimes they boost AOV via free shipping thresholds. Measure it.
Gross profit per session. The most important metric. A promotion that adds 20 percent more orders but cuts margin by 25 percent is a loss. Compute (revenue minus discount minus COGS) per session for both groups.

Step 5: read the result with the right baseline

The lift is treatment minus control during the same period, not treatment now versus treatment last week. Same period because seasonality and macro conditions affect both groups; same period also because external events (a competitor launching a sale, a press feature, a Google algorithm update) affect both groups.

Most merchants who run this test for the first time discover their promotions deliver less lift than expected. Tinuiti’s published benchmarks suggest realistic lift on conversion rate is 5 to 15 percent for percent-off promotions and 8 to 25 percent for free shipping promotions, with gross profit lift typically half that, since the discount eats into margin. Anything above those ranges deserves a second look for measurement error.

Cross-channel coordination: how the three features interact with paid Shopping and Performance Max

The three features sit alongside paid Shopping and Performance Max campaigns rather than replacing them. Understanding the interactions prevents budget waste and avoids accidental cannibalization.

Free listings and Performance Max

Performance Max campaigns ingest the same Merchant Center feed used for free listings. A SKU disapproved for free listings is also unavailable for Performance Max Shopping inventory. Improving feed quality benefits both channels at once. Tinuiti’s Performance Max audits, summarized in their Shopping benchmark series, repeatedly highlight feed cleanup as the single highest-ROI activity for accounts running Performance Max.

The reverse also holds: a Performance Max campaign that bids aggressively on a category will not push out the corresponding free listings. The two surfaces serve different placements and different user intents. The free listing on the Shopping tab below the paid carousel still earns clicks, especially from users scrolling past the paid section. Search Engine Journal documented a mid-2023 case study with an apparel retailer where free listings represented 22 percent of total Shopping clicks even with an aggressive Performance Max campaign running on the same SKUs.

Promotions and Performance Max

Promotion badges appear on Performance Max Shopping placements just as they do on standard Shopping ads and free listings. The badge inherits the same approval rules. Running a promotion across both channels simultaneously is the default behavior; you do not need to set up the promotion twice. The single promotion feed entry covers all surfaces where the SKUs appear.

The interaction worth thinking about: Performance Max bidding takes the badge into account when computing predicted click value. A SKU with a promotion attached typically receives a higher predicted CTR, which can shift the bidding algorithm to be more aggressive on that SKU during the promotion window. Tinuiti’s testing has shown that combining a fresh promotion launch with a Performance Max budget bump often produces a multiplicative effect rather than a simple additive one, since the algorithm leans into the higher-converting inventory.

Automated Discounts and Smart Bidding

Automated Discounts and Smart Bidding (Target ROAS, Maximize Conversion Value) operate on different layers. Smart Bidding sets the bid for the auction. Automated Discounts adjusts the displayed price for a given impression. They do not collide directly, but they share the same underlying conversion-prediction signals.

The cleanest stack for a high-volume account: Smart Bidding on Target ROAS for the paid Shopping or Performance Max campaign, plus Automated Discounts on the same SKU set, plus an active promotion feed for time-bounded offers. Each layer optimizes a different lever. Search Engine Land’s coverage of the Automated Discounts launch quoted Google product managers describing this as the “stacked optimization” model: the paid auction layer, the displayed-price layer, and the promotional metadata layer all running in parallel.

Catalog readiness checklist before turning on the three features

The features compound on top of feed quality. Before turning any of them on, run through this checklist:

Every SKU has a valid GTIN, MPN, or brand plus identifier_exists set to false where the GTIN is not available.
Every SKU has a clean primary image with no overlay text, no watermark, and no border.
Every SKU has a description of at least 500 characters with the most important attributes in the first 145 characters.
Every SKU has a price that matches the landing page price within the tolerance Google allows (typically less than 2 percent variance).
Every SKU has availability set correctly and refreshed at the cadence required by inventory turnover.
The feed contains tax and shipping information for every SKU, either through account-level settings or per-SKU overrides.
The website is verified and claimed in Merchant Center, with a current ssl certificate and no policy violations on the dashboard.
For Automated Discounts: cost_of_goods_sold is populated at the SKU level, and minimum_displayed_price is set where you want a hard floor.
For Promotions: the landing page renders the promotion details correctly, the discount applies at checkout, and the dates declared in the promotion feed match the dates on the website.

Skipping any of these steps shifts the failure mode from feed-level rejection to promotion-level rejection or impression-level no-show. The latter is harder to diagnose because the SKU is technically eligible but not actually performing.

Common mistakes

Treating free listings as a separate program from paid Shopping

They share a feed, a policy stack, and a Quality Score signal. Optimizing one optimizes both. Running them as parallel projects with different teams wastes effort.

Using sale_price for permanent discounts

The sale_price attribute is for time-bounded discounts. Setting it permanently signals to Google that the listed price is fictitious and triggers price suggestion warnings. The clean approach is to set price to the actual selling price and use sale_price only when there is a real, time-bounded reduction.

Submitting promotions without testing the landing page

Half of all promotion rejections trace to the landing page. Before submitting, fetch the landing page in an incognito window, verify the promotion is visible, add the SKU to the cart, apply the code if any, and confirm the discount lands. If you cannot complete that round-trip, the reviewer will not be able to either.

Enabling Automated Discounts on a low-volume catalog

Automated Discounts needs signal. On accounts with fewer than 50 to 100 conversions per month per SKU group, the ML defaults to conservative behavior and rarely adjusts the displayed price. The feature works best on mid-market and enterprise accounts. Smaller merchants get more out of carefully designed manual promotions.

Not declaring COGS

If you turn on Automated Discounts without supplying cost_of_goods_sold, Google falls back to a generic margin assumption that is almost always wrong for your category. Provide COGS at the SKU level. The same attribute also unlocks gross profit reporting in Merchant Center, which is useful even outside the Automated Discounts context.

Confusing the Promotions badge with a price change

The badge does not change the price field. It annotates the tile. Your sale_price is still the source of truth. Merchants who expect the badge to render a strikethrough on the listed price are surprised to find that the price field is unchanged; only the badge is added.

Reusing promotion_id across distinct promotions

Reusing a promotion_id replaces the existing promotion. If you need a new offer, use a new id. Reused ids cause reporting collisions that take days to untangle.

Forgetting timezone offsets on promotion dates

A promotion declared in UTC for a US merchant ends mid-afternoon Pacific time, hours before your business day closes. Always declare the offset explicitly. The Help center recommends matching the offset to your primary store timezone.

Conclusion

Free listings, Promotions, and Automated Discounts are three different layers of the same Merchant Center surface. Free listings give you distribution across Google’s properties at zero per-click cost, conditional on feed quality and policy compliance. Promotions add a metadata layer that earns you a colored badge with measurable CTR lift, conditional on landing-page parity and accurate offer terms. Automated Discounts adds a machine-learning layer that adjusts the displayed price per impression, conditional on COGS data and sufficient conversion volume.

None of the three is a silver bullet. All three reward the same operational discipline: a feed that matches the site, a landing page that honors the offer, dates that respect timezones, and a measurement plan that compares treatment to control rather than this week to last week. Merchants who treat Merchant Center as a single integrated surface, with paid Shopping on top of free listings on top of promotions on top of automated discounts, run an order of magnitude more efficiently than merchants who treat each as a separate project.

The next steps are mechanical. Audit your feed for free listings eligibility. Identify the three to five promotions you actually want to run this quarter and write the long_title and dates before you build the feed. Decide whether your conversion volume justifies enabling Automated Discounts, and if so, populate cost_of_goods_sold first. Then run a controlled test on your first promotion and read the lift against a holdout group. The infrastructure rewards merchants who do the boring work first.

Sources

Google Merchant Center Help: About free listings on Surfaces Across Google.
Google Merchant Center Help: Promotions feed specification and required attributes.
Google Merchant Center Help: Automated Discounts setup and eligibility.
Google Merchant Center Help: Sale price attribute and effective dates.
Search Engine Land: Google opens Shopping tab to free product listings (April 2020 rollout coverage).
Search Engine Land: Coverage of Automated Discounts launch and merchant impact analyses.
Search Engine Journal: Free listings performance benchmarks and merchant case studies.
Tinuiti: Quarterly Shopping benchmark reports covering free listings incrementality and promotion CTR lift.
Store Growers: Promotion feed setup walkthrough and rejection-pattern audit.
Smart Marketer: Promotion compliance audit playbook and landing-page parity checklist.