How to Track Competitor Prices: 6 Methods From Spreadsheets to Full Automation

Intro

Every pricing decision you make is a bet on what your competitors are charging. Most retailers make that bet on stale information.

The problem is not that price tracking is difficult. Opening a competitor’s product page takes ten seconds. The problem is that it does not scale.

A retailer tracking 50 products across 10 competitor sites is maintaining 500 data points. At roughly 30 seconds per check, one complete daily pass is over four hours of work, every day, before anyone has made an actual pricing decision.

That is why manual monitoring quietly collapses beyond a small shortlist, usually without anyone announcing it. The sheet just stops getting updated.

This guide covers six ways to track competitor prices, from a spreadsheet you can build this afternoon to fully automated monitoring.

They are ordered by effort and coverage rather than by quality. The right method depends on how many products you sell and how fast your market moves.

Before you choose a method: decide what you are actually tracking

Skipping this step is the most common reason price tracking projects die. Four decisions, in order:

Which competitors. Not everyone who sells your category. The three to eight retailers your customers actually cross-shop, plus any marketplace sellers who appear on searches for your own brand. A list of 30 competitors guarantees you will track none of them well.

Which products. Your revenue is not evenly spread, and neither should your monitoring be. Start with the products that drive margin, the ones customers price-check hardest, and anything where you have recently lost sales without an obvious reason.

Which attributes. Price alone is misleading. A competitor who is 5% cheaper but out of stock is not undercutting you. They are sending you customers. At minimum, capture price, availability and delivery cost, because the number that decides a purchase is the total landed cost.

How often. Daily is a sensible default for most catalogs. Hourly matters only for fast-moving marketplace listings. Weekly is fine for slow categories, and honest weekly data beats aspirational daily data that nobody collects.

Write these four answers down before you evaluate a single tool. They determine which of the following methods is even viable for you.

Method 1: Manual checks in a structured spreadsheet

The starting point, and genuinely the right answer for small catalogs.

Build one row per product and one column per competitor, with your own price in the first column. Add a date column and a stock column, the two fields people leave out and later wish they had.

Check at a consistent time of day. Prices move intraday, and a Monday-morning number compared against a Friday-evening number is not a comparison at all.

Two refinements make manual tracking substantially more useful:

• Record the gap, not just the price. A column calculating your price as a percentage of the cheapest competitor turns a wall of numbers into something you can sort and act on.
• Keep the history. Do not overwrite last week’s figures. Copy the sheet or append rows. A single snapshot tells you where you stand; a series tells you who is discounting, how deeply and how often.

Realistic ceiling: around 20 to 30 products across three or four competitors, checked weekly. Past that, the arithmetic above takes over.

Method 2: Price alerts and change detection

Instead of checking pages on a schedule, have something watch them and tell you when a number changes.

General-purpose website change-detection services such as Visualping monitor a URL and notify you when the page differs. Browser extensions and price-history trackers do the same for major marketplaces. Setup takes minutes and the cost is often nothing.

This is excellent for one specific job: watching a small number of critical products where any competitor move demands a same-day response.

It is poor at everything else. You get notifications, not data: no history to analyse, no view across your catalog, and no way to answer “what happened to our category last month”.

Watch more than a handful of pages and alert fatigue sets in, at which point you start ignoring the emails and are effectively back to no monitoring at all.

Use it as: a supplement to another method, never as your system.

Method 3: The free competitive data you already have

This one is consistently underused. If you advertise or sell on the major platforms, you are already sitting on competitive pricing signals you have probably never opened.

Google Merchant Center includes price competitiveness reporting that benchmarks your prices against other retailers advertising the same products, using the product feed you already submit.

Marketplace seller dashboards expose comparable signals: buy-box competitiveness, featured-offer pricing, and automated pricing rules that react to competing offers on the same listing.

Be clear about what this is, though, because it is easy to mistake for a complete solution.

It is a thermometer, not a diagnosis. It tells you that you are running warm. It cannot tell you why, or what to do about it.

Specifically, this data cannot give you:

• Per-competitor detail. You learn you are 8% above a benchmark, not which retailer is undercutting you on which product.
• Coverage of independent webshops. It only sees the platform you are on, and many of your real competitors sell elsewhere.
• Coverage of products you do not list there. Anything outside that channel is invisible.
• Usable history, alerting, or stock and delivery context.
• Any MAP visibility at all, if you are a brand rather than a retailer.

Use it as: a free reality check when you have no monitoring in place. It is clearly better than nothing, and often the fastest way to prove internally that the problem is real.

What it cannot do is replace the granularity of a tool built for the job: per-competitor prices on the specific products you care about, across independent webshops as well as marketplaces, with history, stock and alerting attached.

Platform reports tell you that something is wrong. A dedicated tool like Altosight tells you what, where and by how much.

Start here if you have nothing, but do not mistake the benchmark for the picture.

Method 4: Semi-automated spreadsheets

The natural next step for anyone comfortable in Sheets is to let the spreadsheet fetch the prices itself.

Google Sheets can pull data from a page with IMPORTXML, using an XPath expression to target the price element:

=IMPORTXML("https://competitor.com/product-page", "//span[@class='price']")

Set expectations before you invest a day in this: on modern e-commerce sites, it will fail more often than it works.

Most current storefronts render pricing in the browser after the initial page load, and IMPORTXML only sees the raw HTML that arrives first, so the cell comes back empty.

Test it against ten of your actual competitor URLs before building anything on top of it. If three of them return a price, that is a normal result.

The other failure modes matter just as much:

• Site redesigns break every formula silently. A changed CSS class turns your price column into errors, or worse, leaves a stale number that nobody notices.
• There is no anti-bot handling. Repeated automated requests from one source get rate-limited or blocked, often intermittently, which produces gaps rather than obvious failures.
• Nothing tells you it has stopped working. A manual sheet that stops being updated is visibly abandoned; an automated sheet that quietly stops updating still looks authoritative, and people price against it.

Realistic ceiling: a modest number of simple, static product pages, with someone willing to repair formulas whenever they break. Useful as an experiment, fragile as infrastructure.

Method 5: Building your own scraper

With engineering resource, you can build exactly what you want: a script that fetches competitor pages on a schedule, parses out price and stock, and writes to your own database.

For static pages this is genuinely straightforward, perhaps a few dozen lines of Python. The difficulty is everything around it:

• Rendering pages that build their content in the browser
• Handling geo-targeting, currency and locale variants so you are comparing the right market
• Rotating requests responsibly and surviving anti-bot measures
• Retrying failures, and distinguishing “the price is unavailable” from “our scraper broke”
• Scheduling, storage, alerting and reporting on top of the raw data

None of these is hard in isolation. Together they are a product, not a script.

The real cost is not the build but the maintenance. Every competitor redesign breaks a parser, and parsers break silently. Teams routinely underestimate this by an order of magnitude.

Sensible when: you have engineering capacity to spare, unusual data sources no vendor covers, or a hard requirement to keep raw data inside your own systems.

Method 6: A dedicated price monitoring platform

At a certain catalog size the question stops being “how do we fetch these prices” and becomes “how do we keep thousands of comparisons accurate without hiring someone to babysit them”. That is where dedicated software earns its cost.

What you are paying for is not the scraping. It is the crawling infrastructure, the product matching, the normalisation, the price history, the alerting, and the connection into whatever you do next: a repricing rule, an enforcement workflow, a report for your buying team.

Start with one question: do you supply the URLs, or does the tool find the products?

This is the first thing to ask and the most commonly skipped.

URL-based monitoring means you paste in a competitor product URL for every comparison you want. That is precise, but the setup work is yours, and it breaks when a competitor changes their URLs.

Channel-based monitoring means you nominate a competitor site or marketplace and the platform finds and matches your products on it. That is the only approach that stays manageable across a large catalog.

Hybrid setups combine both, and the better platforms support every model, letting you pin exact URLs where precision matters and leave discovery to the platform everywhere else.

Whichever tool you shortlist for competitor price monitoring, ask which of the three models it runs before you start a trial. The answer decides how much of the ongoing setup work lands on your team rather than the vendor’s.

Then work through the rest

• Can it track any website, or only marketplaces? Plenty of tools handle the big marketplaces well and independent webshops badly. Your competitors probably include both.
• How does it match products without identifiers? See the next section. This is the single most important question and the one most often glossed over.
• Which countries and currencies? If you sell cross-border, monitoring that assumes one market is not monitoring.
• Can you get your data out? An API and clean exports decide whether the tool integrates with your pricing process or becomes another dashboard nobody opens.
• Does it support what you actually do with the data? Automated repricing with hard margin floors if you are a retailer; violation detection and documentation if you are a brand enforcing MAP.

The part everyone underestimates: product matching

Every method above assumes you already know which competitor page corresponds to which of your products. In practice, that assumption is where price tracking actually breaks.

The same product appears under different titles, different bundles and different variant structures across retailers. Identifiers are missing, wrong or reused.

One retailer sells the single unit, another sells a two-pack, a third bundles an accessory. Without reliable identifiers, a significant share of listings need disambiguation before their prices can be compared at all.

This is why describing the work as “a 500-URL spreadsheet” is misleading. Fetching 500 prices is the easy half.

Knowing those 500 prices refer to the products you think they do, and keeping that true as competitors restructure their catalogs, is the half that consumes the time.

It is also why matching accuracy, and how it is verified, should dominate your evaluation of any automated approach.

How often should you actually check?

Daily monitoring suits most catalogs. Increase frequency for fast-moving marketplace listings, where competing offers can change several times a day, and for the small set of products where being wrong is expensive.

Plan for peak season separately. Whatever cadence feels comfortable in summer will feel slow during Black Friday week, when competitors reprice within the day and a once-daily snapshot misses the moves entirely.

If you only upgrade your monitoring once a year, do it in October.

What to do with the data, and what not to do

The most expensive mistake in competitor price tracking is treating the lowest observed price as an instruction.

A meaningful share of competitor price drops are stock-clearance events, feed errors or plain mistakes. Automated blind price-matching transmits those mistakes across an entire market within hours, and everyone’s margin absorbs the cost of one retailer’s typo.

Any automated response needs guardrails: hard price floors, margin rules, and a sanity check that ignores implausible drops rather than chasing them.

It is also worth testing the assumption that you must always be cheapest. Availability, delivery speed and trust often carry more weight than a small price gap, and “cheapest at all costs” is a strategy for pure commodity products rather than a default setting.

Good monitoring should answer three questions:

• Where am I priced out of the market?
• Where am I leaving margin on the table?
• What changed this week that I need to know about?

If your setup cannot answer those three, more data will not help.

Which method fits you

Situation	Start with
Under 30 products, few competitors	Structured spreadsheet (Method 1)
A handful of critical products to watch	Change-detection alerts (Method 2)
Already advertising or selling on major platforms	Free platform reports (Method 3)
A few simple product pages, no budget	Semi-automated Sheets (Method 4)
Unusual sources and spare engineering capacity	Custom scraper (Method 5)
Hundreds of products, or matching without identifiers	Dedicated platform (Method 6)

Most retailers move up this list in order, and there is nothing wrong with that.

The failure mode is not starting too small. It is staying on a method long after it has stopped producing trustworthy numbers, and continuing to price against it.