Your rank tracker reported a good week. Position three held across eleven of your top terms, nothing dropped more than a place, and the Monday deck more or less wrote itself. On those same queries, a competitor you don’t monitor took four Merchant Listings slots above your organic result, and an AI Overview recommended somebody else on the strength of a Reddit thread.
SERP competitor analysis is the practice of identifying which brands appear on the search results pages you compete on, feature by feature, and how much of the page each one occupies. That definition doesn’t mention organic rankings, which is where most competitive analysis still begins and ends.
Across 85,465 retail SERP captures in GrowByData’s Retail Industry Panel between July 5 and August 4, 2026, Merchant Listings appeared on 84.35% of pages, against 99.99% for organic. A competitive view built on rank tracking data reads one of those two surfaces and skips the other on roughly five pages in six.
Rank tracking measures your position among ten organic links. SERP competitor analysis measures who occupies the whole page, including the blocks sitting above those links. In retail, those two competitor sets overlap less every quarter.
What does SERP competitor analysis actually measure?
Position is where a domain sits inside a given result set. Presence is whether a domain appears in a given SERP feature at all. Share is what proportion of the listings inside that feature belong to one domain. A standard rank report covers the first of those.
The gap has practical consequences. A brand can hold organic position two and still lose the page, because the eight listings above the fold sit inside a Merchant Listings block it never entered. Its rank report will read as healthy for months while revenue quietly moves elsewhere.
The 84% problem
Every retail SERP capture in the window, scored on whether each feature appeared at all:
Source: GrowByData Retail Industry Panel. 85,465 SERP captures, United States, July 5 to August 4, 2026. Broad retail across home furnishings, apparel, beauty and personal care, food and beverage, toys, consumer electronics, and office equipment.
Merchant Listings against Text Ads is the row pair worth staring at. Most enterprise retail teams have a paid search manager watching text ad competitors daily, sometimes hourly through Q4. Text ads showed up on roughly one retail SERP in ten. The organic shopping surface, which in most organizations has nobody’s name against it, showed up on five in six.
Which makes this a staffing conversation before it becomes an SEO one. When we walk enterprise teams through their own version of this table, the argument that follows is usually between search and merchandising over who owns the product feed, and it tends to run longer than the analysis did.
What does your category’s feature table look like?
Run your domain through the Revenue Risk Report and see which SERP features your competitors hold on your keywords, and which ones you never enter.
Not in our database yet? Talk to an expert and we’ll build the view manually.
Featured snippets are the most overrated target in retail SEO
Five hundred and thirty-eight captures out of 85,465 contained a featured snippet, which works out to 0.63%. Position zero still has a line in most enterprise content briefs and a slide in most agency pitches, and in retail it has quietly stopped existing.
The likely cause sits two rows above it in the table. AI Overviews now occupy the space featured snippets used to hold, on three quarters of pages, and Google has little reason to render both.
A content team writing 40-word answer paragraphs for commercial retail queries is optimizing for something that fires on one page in a hundred and sixty. Moving that effort to AI Overview citation means different mechanics and a measurement problem you’ll have to solve, but at least the target appears. Our note on SERP feature optimization covers where the work pays back by feature.
Your biggest SERP competitor might not be a retailer
Same panel, same window, this time measuring how often two non-retail domains turned up inside each feature.
Source: GrowByData Retail Industry Panel, United States, July 5 to August 4, 2026. Read as: of all captures where the feature appeared, the percentage that included at least one listing from that domain.
Reddit appears in the organic results on more than a quarter of retail SERPs and gets cited in half of all AI Overviews. YouTube is cited in seven out of ten. Neither one sells the product, and both are taking the answer position on queries your merchandising team assumes are purely commercial.
A competitor list built from a rank tracker gives you five retailers and a marketplace. A list built from the page gives you those six plus a forum, a video platform, and whichever aggregator the AI Overview happened to trust that week. The second list explains your traffic better, and producing it is what Google AI Overviews monitoring is for.
Why don’t standard SEO tools show this?
Semrush, Ahrefs, and BrightEdge were built around the keyword-to-ranking relationship, back when a SERP was ten links and an ad block, and they still do that job well. They’ll tell you a page slipped from four to seven and give you a defensible estimate of what it cost. Several now flag whether a feature appeared on the page, which is a checkbox rather than a competitor set.
What they generally won’t give you: which eight sellers held the Merchant Listings carousel on your term in Chicago on mobile yesterday, which domains the AI Overview cited and in what order, or how that composition shifted week over week by product category. Answering those means capturing and parsing the full rendered page rather than reading the ranking positions inside it, which is a different collection problem and a considerably more expensive one.
So, the unvarnished version of the GrowByData pitch: keep your rank tracker. We read the rest of the page. If your category’s competition genuinely does live in the ten blue links, you can skip us.
How do you run a SERP competitor analysis across features?
Five steps, and the order matters, because most teams start at step four and struggle to work backwards from there.
- Build the keyword set from revenue rather than volume. Forty mid-tail terms mapped to your top-margin categories will teach you more than one 90,000-volume head term you have no realistic path to winning.
- Capture the whole page daily, in your chosen markets and devices. National desktop averages hide the thing you need. The feature mix on a mobile SERP in Phoenix isn’t the mix on a desktop SERP in Boston, and the competitor set moves with it.
- Score presence before position. For each feature, answer the binary question first: are you in it at all? A block you never enter usually costs more than a rank that slipped two places, and it’s often cheaper to fix.
- Segment competitors by type. Direct retailers, marketplaces, publishers and aggregators, and community or video platforms behave differently and call for different responses. One undifferentiated competitor bucket produces a chart nobody can act on.
- Give every feature an owner. Merchant Listings belong to whoever controls the feed. AI Overview citation sits with content. Text Ads sit with paid. A feature with no name against it will not improve, however good the dashboard looks.
Step two is where most in-house attempts stall. Capturing full SERP composition daily across a real keyword set is an infrastructure problem rather than an analysis problem, which is why teams tend to buy it instead of building it. Enterprise SERP tracking and Google Shopping monitoring are the two pieces that make the rest of the list possible.
Where this breaks down
Feature-level competitive data produces far more signal than most teams have the process to absorb. One implementation I watched died because the weekly report went from a single ranking chart to fourteen feature charts and the VP simply stopped opening it. If you can’t name the owner and the likely action for a feature, leave it off the report until you can.
Some features also aren’t winnable in any useful timeframe. Discussion and Forums is 94% Reddit. You will not take that block, and a quarter spent trying is a quarter gone. The number still earns its place, because it tells you the category is community driven and that your review and content strategy should account for it.
One caveat on the panel above: it’s retail. Feature mix shifts by category and shifts hard, with non-retail SERPs carrying much lower Merchant Listings presence and higher AI Overview coverage. Read the table as a shape. Your own keyword set is the only benchmark worth quoting internally.
What changes on Monday
Pull your twenty highest-revenue non-brand keywords. Open each live SERP on mobile and write down every domain that appears anywhere above your organic listing, including carousels, AI Overview citations, video blocks, and forum modules.
Compare that against the competitor list in your rank tracker. The delta is your real competitive gap, and it’s the slide worth putting in front of a VP. Twenty keywords by hand takes an afternoon. Four hundred keywords across your chosen markets and devices, refreshed daily, is the part nobody does by hand.
SERP competitor analysis questions we get asked
Is SERP competitor analysis the same as competitive keyword research?
No. Competitive keyword research asks which terms a competitor ranks for. SERP competitor analysis asks who occupies the page on the terms you already care about. The first is a targeting exercise you run quarterly. The second is a monitoring exercise that needs to run daily, because page composition changes faster than keyword portfolios do.
How often should you run it?
Capture daily, review weekly, act monthly. Daily capture matters because feature presence is volatile and a weekly snapshot will miss a block that appeared on Tuesday and left on Friday. Reviewing daily wastes a senior analyst’s time and tends to produce reactions to noise.
How many competitors should you track?
Most enterprise teams pick eight to twelve brands the CMO recognizes and monitor those. Inverting the practice works better: let the page tell you who’s there, then group the results by type. The domains that surprise you are usually the ones worth a meeting.
Does any of this apply outside retail?
The method transfers, though the numbers won’t. Merchant Listings barely appear on non-commercial SERPs, while AI Overview presence runs higher in categories like healthcare, finance, and B2B software. Run the same five steps against your own keyword set and expect a different table.
See the full page, not the ten blue links
GrowByData tracks organic, Shopping, Text Ads, AI Overviews, and 20+ SERP features on your keyword set, by category, market, and device. Book a session and we’ll build the feature table for your categories before you commit to anything.
GrowByData is a search and AI visibility intelligence platform for enterprise brands, retailers, and agencies. We track competitive visibility across Google organic, paid, Shopping, and AI answers in a single view.
