Competitive price monitoring is the repeatable collection of competitor prices for products you have deliberately matched to your own, stored with their history, and used as evidence for pricing decisions. Collecting the numbers is routine engineering. Deciding that their listing and your listing are the same commercial offer is the part that determines whether any of it is worth reading.
Most teams buy monitoring expecting a price feed and discover they have bought a matching problem. Better to know that in week one than in month four.
What it is not
Four things get sold as price monitoring and are not price monitoring.
- A scraper. A scraper returns the number on a page today. Monitoring means the same number, for the same matched pair, on a schedule, with history kept and failures visible.
- A price comparison portal. Portals show what a consumer sees. You need what your buyer sees, including delivery terms, quantity breaks and the VAT treatment in each country you sell into.
- Automatic repricing. Collecting evidence and acting on it are separate systems with separate risks. Plenty of businesses should do the first and never do the second.
- A route to being cheapest. If the only output is an instruction to undercut, you have built an expensive mechanism for giving away margin.
Why matching is the harder half
A price only means something next to a price for the same thing. In practice, the same thing is rarely obvious.
Three failure modes cover most of it: the identifier looks identical but the pack size differs, the identifier differs but the product is identical, and the listing is actually a bundle. A 0.75 litre bottle and a 1.5 litre magnum share a producer, a vintage and most of a title. A construction fixing sold in a box of 100 and the same fixing sold loose share an article number at the manufacturer. Neither pair should ever be compared on headline price.
Manufacturer part numbers help and are not sufficient. Retailers mistype them, pad them with their own prefixes, or reuse one number across a product family. Treat a matched pair as a claim somebody can inspect, not as a fact the system produced.
What one observation should contain
A useful monitored price is a small record, not a number.
- The price shown, and the currency
- Whether tax is included, and at what rate
- Delivery cost at a stated basket value, because free delivery over 50 euros is a price change for everything under 50 euros
- Minimum order quantity and the unit the price refers to
- Availability, since an out of stock price is not an offer
- The timestamp, and a copy of the evidence you read it from
How often is often enough
Daily suits most catalogues. Hourly suits a few hundred products where competitors genuinely move intraday, typically marketplace listings and travel inventory.
Work out the cost before choosing. A catalogue of 5,000 products tracked against 4 competitors is 20,000 page reads per cycle. Run daily, that is 20,000 reads a day. Run hourly, it is 480,000. The second number changes which competitors will tolerate you, and it changes your bill.
What good monitoring is actually for
Our position is unpopular but survives contact with data: most pricing mistakes are prices set too low, not too high.
Work the arithmetic. An item sells at 100 euros and costs 70, so gross profit is 30 euros per unit. Drop the price to 95 and gross profit falls to 25. To earn the same total profit you now need 30 divided by 25, which is 1.2, so 20 per cent more units. A five per cent price cut demands a twenty per cent volume gain. Very few five per cent cuts deliver that.
The same data that supports a cut also shows where you sit 8 per cent below the cheapest credible competitor for a reason nobody remembers. Those findings pay for the exercise.
How the unit of comparison changes by sector
The mechanics are constant. What changes is what counts as one comparable thing.
- Retail: the SKU, complicated by bundles and pack sizes
- Hotels: the room type on a given date with a given cancellation policy, which is three variables, not one
- Construction: the article, complicated by trade discounts that never appear on a public page
- Pharmacy: the pack, where regulated and unregulated lines need separate treatment
- B2B distribution: the contract tier, where the published list price is often the least relevant number on the page
A reasonable first ninety days
Pick 100 products that carry real margin. Match them by hand against 3 to 5 competitors. Collect daily. Spend the first month reading the data and correcting matches rather than changing prices. If a third of the matches need correcting, that is normal, and it is a strong argument for reviewing matches before automating anything.
Tools such as PriceRoom exist to run that loop across markets in Europe, the Middle East and North America, but the discipline matters more than the tool: a small matched set, verified pairs, history kept, decisions written down.