Competitor price monitoring should help a UK retailer or wholesaler spot a meaningful change. A lower number is not enough if it belongs to a different pack, excludes delivery or comes from a listing that cannot be ordered.
Start by agreeing what makes two offers comparable. Then separate a genuine price movement from a changed product match, unavailable stock or a failed check. Here is a practical way to design that review, including a fictional example you can calculate by hand.
Are we comparing the same product and offer?
Keep a maintained link between your product and each competitor listing. Check the manufacturer, model or product identifier, variant, condition, pack quantity and included accessories. A similar title or photograph is a clue to investigate, not a confirmed match.
Treat a different pack as a separate offer, even when the individual items appear identical. You may be able to compare a price per item, litre or kilogram, but retain the original quantity and any minimum order. A unit-price calculation does not make different purchase requirements disappear.
Google's product data specification distinguishes product identifiers, variants, price, availability, unit measures and delivery. It is a useful reference for the fields a comparison needs; it does not mean Google has validated a competitor match for you. Put an ambiguous listing into review before using it in an alert.
Worked example: is the £20 pack actually cheaper?
This is a fictional calculation, not a customer result. Assume the individual items are identical and all prices use the same VAT basis, currency and delivery destination. Your offer is ten items for £25 including delivery. Competitor A lists eight items for £20, with £4 delivery. Competitor B lists ten items for £22 including delivery, but is out of stock.
Your delivered cost is £25 ÷ 10 = £2.50 per item. Competitor A's is (£20 + £4) ÷ 8 = £3 per item. Its lower pack headline does not make it cheaper per delivered item in this example. Competitor B's £2.20 per item is a useful observation, but the unavailable offer belongs in a different category from stock a buyer can order now.
This compares cost per item, not the checkout total for an equal-sized basket. Before acting, check the quantity the customer actually needs, complete packs, order minimums and any delivery threshold or discount. If the VAT basis or delivery charge is unknown, mark the comparison incomplete rather than fill in a convenient zero.
What information should sit beside a price-change alert?
Show the product match, seller, source link, previous valid observation, new observation and the time each was checked. Keep currency, VAT basis, pack quantity, stock and the relevant delivery assumptions visible. A reviewer should be able to understand the change without starting the research again.
Distinguish a price movement from a change in the comparison. If the listing now selects a different size, the delivery destination changes or a promotion ends, explain that difference. Decide which changes deserve an immediate alert and which can wait for a daily review. There is no useful universal percentage threshold for every catalogue.
Avoid sending the same unchanged exception on every scan. Keep its review status and the evidence behind it, then reopen it when something relevant changes. A resolved alert and a successful collection are different events.
What happens when a page changes or a check fails?
A missing price is not a free product. A failed stock check is not proof that the item is unavailable. Preserve the last valid observation with its original time, mark the current check as failed or uncertain and keep it out of a fresh-price comparison until it has been checked.
Our completed competitor monitoring project recorded price, stock and scan-health history and protected the previous valid result when a scan failed. That operational distinction is useful: the business needs to know both what was last observed and whether it is still current. It does not establish that every new website or data source can be connected.
For a new build, check the permitted source and available access before agreeing collection. Use an authorised feed or API where suitable, and define what happens if access or the page layout changes. Give the team a manual route: open the source, confirm the offer and record a dated observation or a match correction. Keep that correction visible to the next automated run.
Does an alert mean we should change our own selling price?
No. It is evidence for a commercial review. The team still needs to consider its own costs, margin, availability, customer offer and service. An unavailable competitor listing or temporary promotion may deserve investigation without justifying a price cut.
The competitor price monitoring demonstration shows prepared prices, stock and movements with fictional records. Its refresh does not visit competitor websites, and it does not publish a new selling price. It illustrates a review experience rather than a live collection service.
If a later workflow will update your prices, agree that as a separate scope. Our price-change approval demonstration shows individual and batch review using fictional data. A recorded approval is still separate from proof that a production website received the right value.
What should a first pilot prove?
Choose a small product set your team can check manually. Include a different pack size, an unavailable offer, a changed variant, unknown delivery and an unavailable source. Agree the expected outcome before running the pilot. The job is to prove that the workflow separates a useful comparison from a reason to investigate.
Measure the work spent gathering and correcting observations, the product matches a reviewer accepts and whether alerts lead to a useful review. Count the checking that remains. More frequent scans or more notifications are not, by themselves, evidence of a better commercial result.
Keep one owner for the product matches and the exception queue. Agree a collection frequency the team can use, who responds to a failed check and how the manual process continues if the automation stops. Once that small set is dependable, decide whether expanding it would return useful capacity.
Eight-item pack versus ten-item pack: Keep both quantities, calculate on an agreed unit and delivery basis, and review the actual basket before a pricing decision.
Cheaper offer marked out of stock: Show it separately from available offers; do not treat it as a directly purchasable alternative.
Missing delivery charge or unclear VAT basis: Mark the comparison incomplete; do not assume a zero charge or identical tax basis.
Price read fails or returns an unexpected value: Hold the new observation for review and retain the previous valid value with its original time.
Product variant changes: Pause the match for review rather than report a price movement on the wrong item.