Comparison Shopping Engine Analytics: Metrics That Matter
Short answer Track profit first: return on ad spend (ROAS) per product against your break-even ROAS. Then three diagnostics: impression share (are you showing?), CTR against the benchmark CTR (do shoppers pick you?), and your price against Google's price benchmark (are you priced out?). Google Ads, Merchant Center and Microsoft Advertising report all of them free.
Want help applying this to your business? Talk it through on a free 30-minute call.
Book a callThe comparison shopping engine metrics worth tracking
Six numbers, in this order. The first tells you whether to keep spending. The rest tell you what to fix.
| Metric | Question it answers | Where to find it | If it's bad |
|---|---|---|---|
| ROAS (return on ad spend) per product | Is this product making money? | Google Ads Products page; Microsoft product partition report | Lower the bid or exclude the product |
| Impression share | How often do we show when we could? | Google Ads Product groups page; Microsoft campaign and ad group reports | Raise budget or bid on winners |
| CTR vs benchmark CTR | Do shoppers pick us over similar listings? | Product groups page; Microsoft product partition report | Fix title, image or price |
| Max CPC vs benchmark max CPC | Are we bidding far below rivals? | Same as above | Raise bids only where ROAS allows |
| Your price vs price benchmark | Are we priced out? | Merchant Center pricing reports | Reprice, or accept fewer clicks |
| Conversion rate | Do clicks turn into orders? | Google Ads, Microsoft, GA4 | Fix the product page, stock or shipping cost |
For most stores, Google Shopping is the engine that matters, so most of this lives in Google Ads and Merchant Center. Microsoft Shopping (Bing) reports the same metrics under slightly different names. Smaller engines usually report only clicks and spend, so you measure them in your own analytics (last section).
Start with profit: ROAS against break-even
ROAS is revenue divided by ad spend. Microsoft's reporting docs give the formula as "(Revenue / Spend)." A ROAS of 4 means $4 in sales per $1 of ads.
ROAS alone doesn't say if you made money. Compare it to your break-even ROAS: 1 divided by your gross margin. At a 40% margin, break-even is 2.5. Below that, every sale loses money once the ad is paid. Work this out per product, not for the whole store, because a high-margin product can carry a low-margin one and hide the loss. The full margin math is in contribution margin in marketing.
Average order value (revenue divided by orders) moves ROAS too. A bundle that lifts it raises ROAS at the same cost per click.
Revenue only shows up if conversion tracking with order values is set up. Without it, these platforms report clicks you paid for and nothing about sales.
Impression share: are you even showing?
Impression share is the share of possible appearances you actually got. Google defines it as:
"The impressions you've received on the Search Network divided by the estimated number of impressions you were eligible to receive." (Google Ads API: Metrics field reference)
Two related numbers tell you why it's low: impression share lost to budget and lost to rank. Lost to budget on a profitable product means money left on the table. Lost to rank means your bid, or your product data, isn't competitive.
Two limits to know:
- Google says you can't split impression share, benchmark CTR or benchmark CPC by device or click type.
- In Microsoft Shopping, impression share is "only available with the campaign and ad group performance reports," not per product (Microsoft Advertising API docs).
Click share works the same way for clicks. Microsoft's example: if 10 ads were clicked and 3 were yours, your click share is 30%.
Benchmark CTR and benchmark CPC: how you compare to similar products
CTR (click-through rate) is clicks divided by impressions. On its own it's hard to judge. Shopping campaigns give you a yardstick: the benchmark CTR of similar products from other sellers.
Microsoft's docs say what to do with it:
"If the benchmark CTR is significantly higher than the CTR for your product group, you might consider raising your bid for that product group, or improving the product information, particularly product images and titles." (Microsoft Advertising API: ProductPartitionPerformanceReportColumn)
Benchmark max CPC (Microsoft calls it benchmark bid) is how other advertisers bid on similar products. Google describes it as "an indication of how other advertisers are bidding on similar products." Use it as a check, not a target. Bidding up to the benchmark only makes sense if the product's ROAS can absorb the higher cost per click.
Google Ads shows both columns on the Product groups page and in Predefined reports, the report builder under Insights and reports (Google Ads Help).
Price benchmark: are you priced out?
Shoppers on a comparison engine see your price next to everyone else's. Google's price benchmark is the "average click-weighted price for a given product across all merchants who advertise that same product on Shopping ads" (Google Cloud: Merchant Center Price Competitiveness table). Products are matched by GTIN, the barcode number, so products without one get no benchmark.
Read it with your click data. A best seller priced 15% above the benchmark that has lost clicks is worth repricing or promoting. A unique or handmade product has no match and no benchmark, and that's fine.
Product-level data: which products to push and which to fix
Campaign averages hide the story. In Google Ads, the Products page shows impressions, clicks, average CPC and conversions per item, with brand, price and custom labels beside them (Google Ads Help). Sort it four ways:
| Pattern | Likely cause | Action |
|---|---|---|
| Many impressions, few clicks | Weak title or image, or high price | Compare to benchmark CTR and price benchmark |
| Many clicks, no sales | Product page, stock, shipping cost or price shock | Check the page matches the listing |
| Good ROAS, low impression share | Budget or bid too low | Give it more budget |
| Spend with no sales over a long stretch | Wrong product for paid traffic | Exclude it or lower its bid |
Google ties performance to feed quality: it warns that "price mismatches between your feed and landing page, or incorrect availability, can lead to product disapprovals." A disapproved product shows zero impressions, which looks like a demand problem but isn't. Check Merchant Center for disapprovals before blaming the bid.
One trap: Google says impression counts in item-level reports and campaign-level reports won't match, partly because one Shopping ad slot can show several products and each gets an impression. Compare products with products, not with campaign totals.
Free listings and smaller engines: measure them yourself
Google's free listings show up in Merchant Center Analytics, not Google Ads:
"Free listings are reported in your Merchant Center Analytics under "Organic" traffic." (Google Merchant Center Help: About Summary in Merchant Center Analytics)
Merchant Center's Summary tab splits product clicks into organic and ads. Smaller comparison engines often give you clicks and cost but no sales. For those, add UTM tags (labels at the end of a link) to the product URLs in each engine's feed, so Google Analytics can tell the traffic apart. Google's advice: "When you add parameters to a URL, you should always use utm_source, utm_medium, and utm_campaign." Then compare revenue by source in GA4's Traffic acquisition report. Traffic sources in Google Analytics shows where to look.
Example
A kitchenware store, $3,000 a month on Google Shopping
Overall ROAS is 3.5 against a 30% margin, so break-even is 3.3: barely profitable. The Products page shows cast-iron pans at ROAS 6 with 40% impression share lost to budget, and knife sets at ROAS 1.2. The owner cuts the knife bids and moves $600 to the pans. Same spend; if the pans hold near ROAS 6, more profit.
A bike parts shop with a low CTR
Its CTR on tires is 0.6% against a benchmark CTR of 1.4%. The price benchmark shows its tires $8 above the average. It drops the price on its five top tires and swaps in clearer photos. CTR rises, and the owner checks that ROAS still clears break-even after the price cut.
FAQ
What is a good CTR for Shopping ads?
There's no single good number; it varies by product and price. Use your benchmark CTR, which compares you to similar products.
Is ROAS the same as profit?
No. ROAS counts revenue, not profit. A ROAS of 3 loses money on a product with a 25% margin, whose break-even ROAS is 4.
Why does my product get impressions but no clicks?
Usually the title, the image or the price. Compare the product's CTR with the benchmark CTR, and its price with Google's price benchmark.
Do I need a paid tool for comparison shopping analytics?
Not to start. Google Ads, Merchant Center, Microsoft Advertising and GA4 report every metric here free. Paid tools help most with hundreds of products on many engines.
How often should I check these numbers?
Weekly for spend and ROAS, monthly for price benchmarks and exclusions. Don't judge a product on a handful of clicks; wait for a pattern.
Different for you?
- You run Performance Max, not a Shopping campaign: since June 15, 2026, its product reports count all networks, not just some, so product numbers jumped that day (Google Ads Developer Blog). Don't compare across that date. See Performance Max for a local business.
- You don't know your margin per product: work it out before you judge ROAS. See contribution margin in marketing.
- Clicks come but orders don't: that's a product page problem. See conversion rate formula and /conversion/.
- You can't tell which channel gets credit for a sale: see GA4 attribution.
- You want someone to run the bids and feed: see /paid-ads/.