Price Competitiveness and Best Sellers in Merchant Center 2026: Turning Price Into a Bidding Signal
Two retailers sell the same espresso machine. One runs at 480% ROAS, the other at 190%. Same bidding strategies, same assets, same feed structure. The difference is not in the ad account: the first retailer knows their price sits 4% below the market benchmark, and the second is 11% above it and is finding out right now, from this article.
The price competitiveness report and the best sellers report in Merchant Center are the only place where Google tells you directly how the market values your products and what people are actually buying. Almost nobody uses them. They sit outside Google Ads, they have no dashboard flourish, and they never say “apply this recommendation”. Here is how to read both, how to turn a price gap into a bidding signal, and how to avoid turning the exercise into a race to the bottom.
Price competitiveness: what the benchmark is and where it comes from
The benchmark is a reference market price for a specific product, built from what other retailers charge for that same product in the same country. It is not a plain average across every listing: prices are weighted by clicks, so products that genuinely receive traffic move the benchmark more than a price sitting untouched in someone’s catalogue.
| Field | Meaning |
|---|---|
benchmark_price_micros |
Benchmark price in micros — divide by 1,000,000 |
benchmark_price_currency_code |
Benchmark currency — reconcile with your feed currency |
country_code |
Comparison market; each country has its own benchmark |
| Product identifier | Your offer_id, the join key to feed and Ads reporting |
Three caveats that stop you from misreading it:
- Not every product has a benchmark. It only appears where Google confidently matched your item to other retailers’ listings — which requires correct GTIN, brand and MPN. If half your catalogue has no benchmark, that is a feed quality diagnosis, not a Google limitation.
- It compares the product, not the whole offer. Shipping, warranty, bundles and returns are not in the number. You can be 6% above benchmark with free next-day delivery and still be the better offer.
- The data lags. This is not a real-time feed, so reacting to daily movement is noise-chasing. Weekly and monthly trends are the signal.
The benchmark is not an instruction to cut prices. It is the answer to “why is click-through rate on this SKU half of what it is on the one next to it” once every other explanation has been ruled out.
How to get the report
- Merchant Center UI: Analytics → Price competitiveness. Good for category-level summaries and spotting the worst offenders fast.
- BigQuery export: the Merchant Center transfer gives you daily history and lets you join the benchmark against your own cost of goods. This is the only practical way to evaluate price gaps with margin attached.
- Merchant API: since the Content API for Shopping was sunset in August 2026, every integration lives here. If your exports went quiet, start with the Merchant API migration checklist.
Best sellers: demand, not your shelf
Best sellers answers a different question: what is selling in your category and country, whether or not you stock it. The report exists at two levels.
- Brand level. A ranking of brands within a Google product category. It tells you whose name pulls demand and whether a brand is worth adding at all.
- Product level. A ranking of individual products with relative demand, previous rank, and a flag for whether the item exists in your own feed.
| Signal | How to read it | Decision it drives |
|---|---|---|
| Relative demand | Demand share against the category leader | Distinguishes “one runaway winner” from an even field |
| Previous rank | Position in the prior period | A product climbing from 40th to 12th is a buying candidate |
| In your inventory | Whether you list the item | A ready-made list of assortment gaps |
The single most valuable slice is the set of high-demand products you do not stock. That is a purchasing agenda, and it also explains why a campaign hits a ceiling: you are advertising what you have while the market searches for what you do not.
Turning a price gap into a control signal
Reading the report “for awareness” achieves nothing. The working pattern is to convert the gap between your price and the benchmark into a feed label, then manage bids and structure by that label.
Step 1. Compute the gap
gap = (your price − benchmark) / benchmark × 100%. Positive means you are above market, negative means below.
Step 2. Bucket the catalogue
| Gap | What it means | Action (directional) |
|---|---|---|
| below −10% | Clearly cheaper than the market | Raise ambition: dedicated campaign, more aggressive target — and check you are not selling at a loss |
| −10% to −2% | Healthy competitive position | This is where the budget belongs |
| −2% to +5% | At market | Win on the listing instead: imagery, reviews, delivery, promotions |
| +5% to +15% | Above market; CTR will suffer | Lower the bid or target, strengthen the offer’s value |
| above +15% | Out of market for this product | Pull from ads, change the price, or sell it as a bundle |
Treat those thresholds as a starting point, not gospel. In consumer electronics a 5% gap decides the auction; in niche furniture 15% barely registers. Find your own threshold empirically: plot CTR against gap on your own data and see where the curve bends.
Step 3. Write the buckets into the feed
A gap controls nothing; a label does. Set custom_label_3 = price_below | price_market | price_above and refresh it on a schedule using a supplemental feed or feed rules — no developer required, as covered in feed rules and supplemental feeds. The segmentation mechanics themselves are in custom labels in your product feed.
Step 4. Split buckets across campaigns and targets
From there it is straightforward: price-advantaged products get their own campaign with a more aggressive target, out-of-market products get a cautious target or come out of rotation until the price is reviewed. What you must not do is keep all three buckets inside one campaign on one target — the algorithm averages them and spends on items that cannot win.
Three traps in computing the gap
The formula looks trivial. It is also where most analyses break.
Tax and currency
The benchmark arrives in the currency named in benchmark_price_currency_code, and you must reconcile it against your feed currency, not your Google Ads account currency. The second trap is tax: if one market submits prices inclusive of VAT and another exclusive, the second market’s gap will be systematically overstated. Normalise both sides before drawing conclusions.
Product variants
Where one product exists as five variants — colour, capacity, size — the benchmark is computed per variant. Averaging the gap up to the parent product produces a comfortable-looking number that hides two variants sitting 25% above market and quietly consuming budget.
Sample size
A benchmark built from three competing sellers is far less stable than one built from a mass-market listing. Flag those rows and handle them manually rather than through an automated rule.
A worked example: what closing the gap is worth
Say you list 4,000 products and 620 fall into the “+5% to +15%” bucket. They take 18% of clicks and deliver 9% of conversions — the classic signature of a price problem. CTR on that group runs at, say, 1.1% against 1.9% in the healthy bucket.
Option A — cut bids 30% on the problem bucket. You free budget and redirect it to the healthy bucket. You lose some clicks from the expensive group, but every redirected click converts materially better. Profit improves without touching a single price.
Option B — reprice selectively. Of the 620, perhaps 40 sit at a 6–8% gap with enough margin to absorb a cut. Close those gaps, watch CTR recover towards the average, and return those items to the main pool. The remaining 580 stay on reduced bids until purchasing terms change.
Option C — wait for the algorithm to work it out. It will: it will keep buying those clicks because they are formally relevant to the query, and spend the budget without the sales. Automated bidding cannot see the benchmark. It only sees your conversions after the fact.
Smart bidding learns you are above market through the absence of conversions — that is, after the money is gone. The price competitiveness report gives you the same information beforehand.
Best sellers as a season planning tool
The price report governs current efficiency; best sellers governs the next two to three months. A workable routine:
- Monthly, export the top 100 products and top 20 brands for your categories.
- Compare rank to the previous period. A sharp climb signals a forming trend; by the time an item reaches the top ten, buying it is late.
- Flag items missing from your feed and hand purchasing a monthly list annotated with relative demand.
- Watch brands. A brand you do not carry holding a top-five position for two or three consecutive months is no longer a spike.
- Cross-check against your own data. A high-demand product that you do stock but that receives no impressions is usually a feed or price problem, not a budget one.
Price, margin, and why ROAS lies here
The central trap of price optimisation: cutting price almost always improves ROAS and almost always reduces profit. Revenue rises, margin compresses, the ad account looks better, the business earns less.
So the gap is only meaningful alongside margin:
- Below market, high margin — your core asset. Budget and ambition go here.
- Below market, low margin — the danger zone. Traffic arrives, profit does not. Check whether these SKUs are dragging the whole account down.
- Above market, high margin — work the value side: bundles, warranty, delivery speed, promotions.
- Above market, low margin — a candidate for removal from advertising.
Technically this means bidding on profit rather than revenue — sending gross margin as conversion value instead of order total. How that works and what breaks on the way is covered in POAS vs ROAS. Price benchmark plus profit-based bidding is a rare pairing where automated bidding finally optimises for what the shop exists to do.
What to do when you cannot change price
Plenty of retailers do not control price: it comes from an ERP, a distributor or brand policy. That is not a reason to close the report — it is a reason to look for levers that are not the price tag.
- Promotions. A discount annotation on the listing does much of what a price cut does, without surrendering the price itself — see Merchant Center promotions.
- Shipping. Free or fast delivery shown on the listing frequently outweighs a few percentage points of price gap.
- Ratings. Stars next to a price change how that price is perceived; how to qualify is in seller ratings and product ratings.
- Listing quality. Titles, images and attributes still return more than bid micro-management — see feed diagnostics and product disapprovals.
- Cost per click. If the price will not move, move the cost side: working through a CSS partner changes click economics, as explained in Google CSS for Shopping.
A 30-minute weekly routine
- Monday, 10 minutes. Export benchmarks for your top 100 products by clicks last week. Everything below that has no statistical weight.
- 10 minutes. Recompute gaps and refresh labels in the supplemental feed.
- 5 minutes. Review everything that moved into the “above +15%” bucket — those are next week’s CTR problems.
- 5 minutes. Open best sellers, list the high-demand items you do not carry, and hand it to buying.
- Monthly. Check the trend: is more of the catalogue sitting in the healthy band, and what happened to profit rather than just ROAS?
Five ways this report backfires
1. Matching the benchmark across the whole catalogue
This is the race to the bottom in slow motion: you cut, competitors see a lower benchmark, they cut, the benchmark drops again. Match selectively, on products where price genuinely decides the sale.
2. Reading gaps without click volume
A product with two clicks a month and a +30% gap does not deserve your attention. Sort by volume, not by gap size.
3. Treating a missing benchmark as a Google bug
Usually it means a missing GTIN or a wrong brand value. An empty benchmark is a prompt to fix identifiers — and to check product disapprovals and feed diagnostics while you are there.
4. Ignoring country and currency
Benchmarks are computed per country. Comparing one feed’s price against another market’s benchmark is the most common source of “strange numbers” in multi-market accounts.
5. Blaming price for a reporting change
After the Merchant Center reporting changes, many merchants saw organic impressions fall and concluded their pricing had broken. Those are separate stories — what actually changed is in Merchant Center cross-platform reporting.
How this connects to the rest of the account
Price decides whether you win the click. The landing page decides what happens next — and in automated campaign types Google increasingly chooses that page for you; the controls are in final URL expansion in Performance Max and AI Max. Neither works alone: a keenly priced product with a dead landing page does not sell, and neither does a perfect landing page for something priced 20% above the market.
When price buckets have to be maintained across many accounts and teams, it is far easier from a manager account with shared rules and clean access control — see Google Ads agency account rental. More Shopping and paid traffic material sits in the PPC Rebels blog.
FAQ
Who can access the price competitiveness and best sellers reports?
Merchant Center accounts that meet the minimum eligibility requirements and comply with the Merchant Center terms. In practice that means a working feed with valid product identifiers and enough data volume in the target country.
How is the benchmark calculated?
As a reference market price for the same product across other retailers in the same country, weighted by clicks so that listings receiving real traffic count for more.
Does shipping count towards the benchmark?
No. It compares product price only, which is why an offer with free delivery can win while sitting nominally above the benchmark.
Why do some products have no benchmark?
Google could not match your item to comparable listings — missing GTIN, an incorrect brand or MPN, or a genuinely rare product. Start with feed identifiers.
Can I automate repricing directly from the benchmark?
Technically yes, practically risky: the data lags and blind benchmark-following drives prices down for everyone. Use it as an input to bidding and assortment decisions and change prices deliberately.
How does relative demand differ from rank?
Rank is a position. Relative demand shows the distance to the leader — sometimes first and second are worlds apart, sometimes the whole top 20 is level.
Is best sellers useful in a narrow niche?
Yes, at brand level and through rank movement. Even in a small category the report shows which brand is gaining demand, which is a buying decision for next season.
How often does the data refresh?
Not in real time. Work in weekly slices; daily movement is noise.
How do I join the benchmark to Google Ads data?
By product identifier. The cleanest route is exporting both Merchant Center and Google Ads into BigQuery and joining on offer_id, so price, benchmark, cost, clicks and conversions end up in one table.
Should I remove everything above a 15% gap?
Not automatically. Check whether CTR and conversion rate actually dropped. Unique items and bundle anchors can justify a premium.
Does price influence Shopping ranking directly?
Price is not a standalone ranking factor, but it moves CTR and conversion rate, which in turn affect how often your listing is served. The effect is indirect and strong.
What if the benchmark contradicts my own competitor research?
Check country, currency and product matching first. Usually the discrepancy is scope: you are comparing yourself to ten known competitors while the benchmark reflects the whole market, including sellers you never think about.
