· 15 min read
How to Track a Competitor's Shopify Store: Free Weekly Log
Track a competitor's Shopify store with a free weekly log: six public signals worth recording, how to read the deltas, and what a snapshot can't show.
Most competitor research produces a snapshot: you open a store once, read its product pages, note its apps, and close the tab. That snapshot tells you what the store looks like. It cannot tell you what the store is doing — and doing is where the strategy lives.
This guide is about the other half. Once you have a list of stores worth watching — the store discovery methods cover how to build one — the next move is to stop reading them as pages and start reading them as a time series. One row per store per week, six fields, twenty minutes. It is the cheapest research habit in ecommerce, and almost nobody does it consistently.
Everything below uses only publicly visible information and free tools: the storefront, its sitemap, the Meta Ad Library, and a browser. No paid dashboards, no scraping infrastructure, no guessing at numbers nobody can see.
Why a Snapshot Is the Wrong Unit
A single visit gives you states: this store has 84 products, charges $38 for its hero item, runs 11 ads, uses a premium theme. Every one of those numbers is uninterpretable in isolation. Is 84 products a lot? Is $38 a high price? Is 11 ads a serious budget or a test?
Deltas answer those questions without any external benchmark, because the store becomes its own baseline:
- 84 products means nothing. 84 products after 61 last month means the store added 23 SKUs — a launch wave — which is a decision someone made with money behind it.
- $38 means nothing. $38 down from $45 on the hero SKU, while the bundle price stayed flat, means a margin experiment rather than a clearance.
- 11 ads means nothing. 11 ads after 3 ads for six weeks means someone just decided to scale.
You are not trying to learn the market’s average. You are trying to catch the moments when a competitor changes their mind — because those moments are the only public evidence of what they believe is working.
The Six Fields Worth Logging
The temptation is to record everything. Don’t. Six fields, each readable in under three minutes per store, cover most of the signal:
| Field | Where it comes from | What a change implies |
|---|---|---|
| Product count | Product sitemap, total /products/ URLs | Catalog expansion, purge, or seasonal rotation |
| Basket price | Same 5 SKUs, price shown to a logged-out visitor | Pricing experiment, clearance, or cost pressure |
| Ads active | Meta Ad Library, active ads for that page | Scaling, pausing, or a new creative angle |
| Oldest active ad | Start date on the longest-running creative | Whether the account is stable or churning |
| Review count | Review widget on the 3 hero products | Demand velocity, independent of ad spend |
| Theme and app signature | Page source: theme name, store id, app assets | Reinvestment, replatform, or stack change |
Each of these is a proxy, not a fact. None of them measures sales. That limitation is the point of the next section, but the proxies are strong enough to build a timeline from.
Product count from the sitemap
Shopify publishes a product sitemap whether the merchant thinks about it or not. The full method for counting products from the sitemap covers the URL patterns and the large-store sharding problem; for weekly tracking you only need one habit: count the /products/ URLs the same way every week, and record the number plus the date of the newest lastmod entry you see.
Two cautions. Count product URLs, not <loc> tags — the sitemap file starts with the homepage and may include collection or blog sitemaps. And a store that adds forty variants of one product will not move this number, because variants live inside a product URL, not beside it. Product count tracks catalog breadth, not SKU depth.
Basket price from a fixed list
Pick five products before you start: the hero item, two mid-catalog items, one accessory, one bundle or kit. Write the exact URLs into your sheet and never change them mid-stream, or your price series breaks.
Record the price a logged-out shopper sees. Note whether a compare-at price is displayed, because “$45 struck through, $38 now” and “$38” are different events even though both show 38 in the cart.
Ad activity from the public library
The free Ad Library workflow explains how to read creative age and message testing. For a weekly log, two numbers are enough: how many ads are active right now, and what start date the oldest still-running ad carries.
Ad count is noisy — platforms recycle and relaunch creatives — so treat a single week’s move as weak evidence and three consecutive weeks as a trend. The oldest-ad date is the steadier signal: an account whose longest-running creative is 200 days old is running something that works, while an account that resets to 20 days every month is still searching.
Review volume on hero products
Reviews accumulate slowly and publicly, which makes them a decent demand proxy that is independent of how much the store spends on ads. Log the visible count on the same three products each week.
The number itself is less interesting than its rate relative to the store’s ad activity. Reviews climbing while ad count stays flat suggests organic or repeat demand. Reviews flat while ad count triples suggests the new spend is not converting, or is converting on a different product.
Theme and app signature
Page source exposes the theme name, the theme store id, and asset paths from installed apps. The theme detection guide walks through the three fields and what they mean. For tracking purposes, one line in your sheet is enough: theme name plus any new third-party asset domain you did not see last week.
A theme change is a capital decision — someone paid for a redesign or a migration. A new app asset domain is a smaller but still deliberate decision, usually about email, reviews, subscriptions, or support.
Collection structure
The cheapest field nobody logs: the names of the store’s collections. New collection names are positioning statements. A store that adds “Gifts Under $50” in late October is telling you where it expects Q4 demand. A store that adds “New Arrivals” and stops updating “Sale” is telling you the discount cycle ended.
Build the Watchlist Before You Build the Log
Five to ten stores, not fifty. The constraint is not how many you can find; it is how many you will actually revisit on the same day every week for two months.
Three rules make the list usable:
- Same buyer, not same size. A store selling to your customer at a different scale is comparable. A store selling to a different customer at the same scale is not. Size mismatch changes tactics; customer mismatch changes everything.
- Same platform. Shopify-only keeps the signals comparable — sitemap behaviour, theme ids, app signatures. Mixing platforms means half your fields do not exist for half your list.
- Active now. A dormant store produces a flat line that teaches nothing. Confirm current activity before the first row, using the legitimacy and activity checklist as the filter.
Then fix the calendar. Same weekday, same hour, because ad counts and sale banners swing with the week. Monday morning is a reasonable default: it captures whatever ran over the weekend and lands before the week’s own work starts.
The Twenty-Minute Weekly Loop
The example figures below are hypothetical and illustrative. They are there to show the shape of the loop and the kind of arithmetic the log supports, not to describe any real store.
- Open the sheet, copy last week’s row (2 min). Duplicate the row for each store so this week’s numbers sit beside last week’s. Deltas should be a formula, not mental arithmetic.
- Count products (5 min). Open the product sitemap, count
/products/URLs, note the newestlastmod. (Hypothetical: 61 last week, 84 this week.) - Read the basket (5 min). Check the five fixed URLs in a logged-out browser window, record each price and whether a compare-at is showing. (Hypothetical: hero item $45 → $38, others unchanged.)
- Check ads (4 min). Ad Library, advertiser page, active ad count and oldest start date. (Hypothetical: 3 → 11 active, oldest still 180 days old.)
- Reviews and stack (3 min). Same three products, visible review counts; plus theme name and any new app asset domain.
- Write one sentence of interpretation (1 min). This is the field that makes the log worth keeping, and the one everyone skips.
That last step is not decoration. Raw numbers age badly — six weeks later you will not remember why 84 mattered. A sentence like “23 new SKUs, all in the gift category, hero price cut 15% — looks like a Q4 push” preserves the reasoning, and wrong reasoning you can later correct is worth more than data you cannot interpret.
Reading the Deltas
Most changes are noise. The skill is separating the four or five patterns that mean something from the churn that means nothing.
| What you observe | Plausible reading | What to check next |
|---|---|---|
| Product count up sharply in one week | Launch wave, seasonal drop, or a bulk import | Are the new URLs one collection or scattered across categories? |
| Product count down sharply | Purge, end-of-season, or migration | Do the removed URLs 404, or redirect to a collection? |
| Hero SKU price cut, basket flat | Margin or conversion test on the flagship | Did the bundle price move too? Did ad count rise with it? |
| Basket price cut across all five | Clearance or a demand problem | Is a sitewide banner or discount code now visible? |
| Ad count triples, oldest ad still old | Scaling a proven creative set | Did product count or price move at the same time? |
| Ad count goes to zero | Pause, account reset, or budget stop | Is the store still updating products and reviews? |
| New collection names in a new category | Positioning shift or test | Do the new collections have ads behind them? |
| Theme or app signature changes | Reinvestment or replatform | Did page structure, PDP layout, or review provider change? |
Two patterns deserve special attention because they are the ones most people misread.
The catalog surge that isn’t a launch. A jump from 61 to 84 products looks like a launch wave. But if the new URLs all sit in one collection with near-identical titles, it is more likely a bulk import of variants or a supplier feed sync. Check whether the new products have ads and reviews before concluding the store made a strategic bet.
The ad pause that isn’t a shutdown. Ad count dropping to zero reads as “they stopped” and often means “they paused a flight” or “the account is between tests”. Cross-check the store side: products still updating, reviews still climbing, banner still rotating → still in business, just not buying reach this week.
A Worked Comparison, With Hypothetical Numbers
Two stores, four weeks, same niche. The numbers are invented to demonstrate interpretation.
Store A goes from 61 products to 84 in week two, cuts its hero price from $45 to $38 in week three, and lifts active ads from 3 to 11 in week four. The basket total drops 12 percent. Reviews on the hero product climb 9 in four weeks.
Read together: a coordinated expansion. New SKUs, a sharper entry price, and paid reach switched on — the shape of a store pushing a season. The review velocity says the push is finding buyers.
Store B stays at 40 products all month, holds prices flat, and drops from 14 active ads to 2 in week three while its oldest creative resets from 210 days to 12 days. Reviews move by 2.
Read together: a reset, not a retreat. The catalog is stable and prices are stable, which argues against distress. But the long-running creative disappeared and the account is now testing from scratch — the thing that worked stopped working, or stopped being affordable.
The lesson from the pair is that no single field carries the story. Price alone would have missed both. Product count alone would have flagged A and cleared B. The log’s value comes from the six fields disagreeing with each other in informative ways.
What Outside Signals Cannot Tell You
This is where most competitor tracking goes wrong: it slides from observation into invention. From the outside, you can see catalog, prices, ads, reviews, and stack. You cannot see:
- Inventory. In-stock and sold-out states are visible on a PDP, but depth is not. A store showing “in stock” may have four units.
- Revenue. No public page exposes sales. Any per-store revenue figure you see from a free tool is an extrapolation with error bars wide enough to drive a truck through.
- Margin and cost. Price is visible; landed cost, shipping subsidy, and return rate are not. A price cut that looks like a margin squeeze may be a supplier negotiation.
- Ad spend. Ad count is not budget. Eleven ads can cost less than two, depending on placement, duration, and bidding. The honest way to read competitor budgets explains why the ranges people quote are softer than they look.
- Conversion rate. Without sessions and orders you cannot compute it, and inferring it from review velocity assumes a review rate you do not know.
The discipline is to write the observation and the interpretation in separate columns, and to keep the interpretation labelled as a guess. A log that says “price cut 15 percent, likely a test” is useful. A log that says “price cut because margins are collapsing” is fiction with a timestamp.
When a Weekly Log Stops Being Enough
The manual log has three hard ceilings, and it is worth knowing where they are before you hit them.
Breadth. Ten stores is a habit. Two hundred is a job. The loop does not scale past roughly fifteen stores before the twenty minutes becomes two hours, and the quality of the notes collapses first.
History. Your log starts the week you start logging. You cannot retroactively ask what a store’s catalog looked like six months ago, which means the first two months of any new log are thin — you are building a baseline, not reading a trend.
Alerts. A weekly cadence misses anything that happens and reverses inside seven days: a 48-hour flash sale, a one-week test price, a creative that ran for four days. If the change you care about is short-lived, weekly is the wrong resolution and manual checking cannot fix it.
Past those ceilings the answer is a database with history and change detection, not more discipline. That is the layer TrendTrack’s platform covers: a store index of 1.5M+ Shopify stores with product, traffic, and ad signals you can filter and compare, rather than the ten rows you can maintain by hand. The free version of the workflow above is where you should start, because it teaches you which deltas you actually care about — a question no dashboard answers for you.
Verified 2026 · 20% off Starter, Pro and Business for your first 3 months
If you go that route, the TrendTrack promo code is WINNING - 20% off all plans for the first 3 months, which keeps the first quarter of database access cheap while you decide whether tracking at scale replaces the spreadsheet or just feeds it.
Start With One Row
The barrier to this habit is not skill, it is the first row. Open a sheet, add seven columns — date, store, products, basket total, ads active, oldest ad, reviews — plus a notes column, and fill it for one store today. Next Monday, fill the same row again.
Two weeks in, you will have something no article can give you: a competitor whose behaviour you can compare against itself. Four weeks in, the deltas will start arguing with each other, and that argument is the research.
FAQ
How often should I check a competitor’s Shopify store?
Weekly is the right default for most stores, on the same weekday each time so the comparison stays clean. Move to twice weekly only if you are actively tracking a launch window or a promotion you expect to be short, and accept that the extra cadence costs accuracy in your notes rather than improving it.
Can I track competitor prices on Shopify for free?
Yes, by hand and within limits. Pick a fixed basket of five products, open them in a logged-out browser weekly, and record the price plus whether a compare-at price is displayed. What you cannot do for free is track thousands of SKUs across dozens of stores continuously, since that requires scheduled checks and a price history you did not build yourself.
Why did a competitor’s product count suddenly drop?
The common causes are a seasonal purge, a discontinued line, a migration to a new theme or platform structure, or products being unpublished while staying live at their old URLs. Check whether the missing URLs return a 404 or redirect to a collection page, and whether the drop happened alongside a price or ad change.
Does a competitor’s ad count tell me their budget?
No. Ad count measures how many creatives are live, not how much is being spent behind them, and eleven short-lived ads can cost less than two long-running ones. Use ad count as a direction-of-travel signal across weeks and treat any specific budget figure attributed to a store as an estimate rather than a measurement.
Can I see a competitor’s sales or inventory from outside?
Not reliably. Inventory depth, order volume, revenue, and margin are all internal numbers that no storefront exposes, and the per-store revenue figures sold by free tools are modelled extrapolations. Public proxies like review velocity, catalog churn, and ad activity are the honest substitute, provided you label them as proxies.
What should a competitor tracking spreadsheet contain?
One row per store per week with the date, product count, a fixed-basket price total, active ad count, the oldest active ad’s start date, visible review counts on three hero products, the theme signature, and one sentence of interpretation. The interpretation column matters most, because numbers without the reasoning behind them are unreadable a month later.