VKT Toolkit · September 25, 2026 · 9 min readHow to Compare Product Prices Across Websites in Chrome (2026)
Comparing prices sounds like the easiest thing in online shopping: open four tabs, look at four numbers, pick the smallest one. It stops being easy the moment the four numbers are not measuring the same thing — one includes shipping, one is a refurbished unit, one is a bundle with a case you do not want, and the fourth is in a currency you did not notice.
That gap between "numbers on screen" and "prices you can actually compare" is where most bad purchasing decisions live. Here is how to close it.
Why Cross-Store Comparison Breaks Down
Four things make prices from different stores genuinely hard to line up:
- Stores do not agree on what the product is. Titles are marketing copy. Two listings with the same title can be different generations of the same product.
- The displayed price is not the paid price. Shipping, tax, mandatory fees and coupons all sit outside the big number.
- Prices move. Several large marketplaces reprice constantly, sometimes daily. A comparison captured on Monday can be wrong on Thursday.
- No single site covers everything. The cheapest offer is often at a shop that no comparison engine has ever indexed.
Any method worth using has to answer at least the first two.
Method 1 — Tabs and a Spreadsheet
The default method, and for three or four items it genuinely works. You open the candidates, copy title, price, model and link into a sheet, add a shipping column, and sort.
Where it breaks:
- Copy errors. Hand-transcribed model numbers are wrong often enough to matter, and they are the one field you cannot afford to get wrong.
- No history. The sheet is a snapshot. Next week you start over.
- It does not scale. Past about five items, the overhead of updating the sheet exceeds the value of the comparison, so people quietly stop and go with the first tab.
Use it when the decision is small and one-off. Do not use it for anything you expect to revisit.
Method 2 — Comparison Shopping Engines
Google Shopping and its regional equivalents (Idealo, PriceRunner, Kelkoo and others) solve part of the problem properly: they catalogue products by identifier, so they are comparing the same object rather than the same words.
Their blind spots are structural, not accidental:
- Coverage follows catalogues. Products without a clean GTIN/EAN — handmade goods, obscure parts, bundles, used items — rarely appear.
- Marketplace pricing is often stale or absent. Repricing happens faster than feeds update, and some large sellers simply do not syndicate.
- Shipping and total cost vary by destination. A headline price is rarely the delivered price for your address.
- They are regional. An engine strong in one country may be thin in another.
Treat them as one input, not as the verdict — especially for anything sold mainly by small shops.
Method 3 — Collect Everything Into One Library
The third method flips the order: instead of asking a service to compare for you, you collect the candidates yourself into one place where they can be compared on your terms. The capture step is one hover and one click, and it works on listing grids as well as detail pages, so gathering six candidates takes about as long as opening six tabs.
VKT Item Pin is built for this: each pin records title, price, link, model, site, date and thumbnail; re-pinning the same product adds a price point instead of a duplicate. Around two dozen marketplaces have hand-tuned selectors, and everything else runs through a universal engine — JSON-LD structured data first, then Open Graph tags, then generic selectors, and finally a manual pick mode where you click the element you want.
Premium adds the piece that makes it a real comparison: filters by site, date and price range, CSV export, and a chart that overlays up to six products so you can see which store was consistently cheaper, not just which one is cheapest today.
Capture the Right Fields, Not Just the Price
A comparison is only as good as the fields behind it. Six are worth insisting on:
| Field | What it prevents |
|---|---|
| Model / SKU | Comparing a newer revision against an older one |
| Site | Forgetting where the cheap one was, two weeks later |
| Date | Treating a stale price as current |
| Link | Losing the exact variant you were looking at |
| Title (as written) | Silently "correcting" the listing into something it is not |
| Thumbnail | Confusing two visually similar listings |
Keep the title as the seller wrote it. Rewriting it into your own words feels tidier and destroys the one record of what was actually advertised.
Normalize Before You Compare
Before sorting by price, make the numbers mean the same thing. Five checks, in the order they usually matter:
- Delivered cost. Add shipping and any mandatory fee. A €3 difference disappears fast once one store ships free.
- Tax treatment. Some prices include VAT and some do not; cross-border offers often exclude it plus duty.
- Condition. New, used, refurbished and open-box are different products at different risk levels, whatever the price says.
- Bundle contents. "Includes charger and case" versus "body only" is not a price difference, it is a different offer.
- Variant. Storage size, colour and region code all move the price without changing the title.
Write the normalized number next to the captured one, or you will re-do this work next week.
Compare Over Time, Not Just Today
A single snapshot answers "who is cheapest right now". Most real purchases want a different question: "is this a good time to buy?" That needs a series.
Two patterns are worth capturing:
- The baseline before a sale. Pin a few weeks ahead of a known sale window, then pin during it. If the line did not move, the discount was cosmetic.
- The cross-store spread. Pin the same model at two or three stores on the same day. If one is consistently cheaper by a fixed margin, that margin is probably structural — and worth more than any one-day promotion.
With a per-item history, "cheapest" stops being a moment and becomes a pattern you can act on.
When a Site Refuses to Be Read
Extraction fails in predictable ways, and each has a workaround:
- No structured data. Smaller shops often skip JSON-LD. The universal engine falls back to Open Graph and generic selectors; if the result is wrong, edit the field before saving.
- Price rendered by script after load. Wait for the price to appear, then pin. Pinning too early captures a placeholder.
- Everything fails. Use pick mode: click the element on the page and the value goes into the field you chose. Slower, but it works on literally any page.
- Login walls and regional redirects. Pin what is actually displayed to you — including your local currency — rather than what a comparison site claims the price is.
Be honest about the last one: a local, logged-in price is the only one you can actually pay, which makes it more useful than an idealized global figure.
The Three Methods at a Glance
| Method | Best for | Breaks when |
|---|---|---|
| Tabs + spreadsheet | One-off decisions with 3–4 candidates | You need to revisit it, or there are more than five items |
| Comparison engine | Catalogued products with clean identifiers | The item is handmade, used, bundled or sold by a small shop |
| Local library | Cross-store comparison you want to keep and update | You need automatic re-checking or price-drop alerts |
The honest limits of the library method: nothing re-checks prices while you are away, so you must re-pin to refresh a number; the free tier holds 50 products; the data lives in one browser profile; and clearing browser data erases it unless you export a JSON backup. Those are the costs of keeping it local.
Frequently Asked Questions
Can I compare prices from stores in different countries?
You can, but the raw numbers are not comparable: currency, VAT treatment, import duty and shipping all differ. A capture tool records exactly what the page displayed, so note the currency and whether tax was included, then convert before deciding.
Does this work on sites other than Amazon?
Yes. Around two dozen marketplaces have hand-tuned selectors, and every other shop goes through a universal engine that reads JSON-LD structured data, Open Graph tags and generic selectors. If all of that fails, you can click any element on the page to fill a field manually.
How many products can I compare at once?
The free tier stores up to 50 products and shows a price-history chart for each one individually. The Premium comparison view overlays up to six products in a single chart, which is the closest thing to a side-by-side comparison.
Do I have to open every product page to capture it?
No. The pin button appears on listing and search-result grids too, so you can capture several candidates from one results page without opening each product detail page in its own tab.
What if the store shows the price in another currency?
The record keeps the number and currency exactly as displayed, with no automatic conversion. That is deliberate: converting silently would hide the difference between a domestic price with tax and an import price without it. Convert it yourself when you compare.
Bottom line: use a comparison engine to discover candidates, then keep your own library of the ones that matter — model number, site, date and delivered cost captured together. VKT Item Pin does the collecting: hover, pin, re-pin to add price points, 50 products free, with a six-product overlay chart on Premium.
More VKT tools live in the extensions catalog, or reach us at [email protected].