Most small retailers and distributors already do price monitoring. It just happens manually, on a phone, in the evening, by someone opening four competitor sites and typing numbers into a spreadsheet. It works until the catalogue passes a few hundred items, and then it quietly stops happening at all.

Automating it is not a large project. The mistakes, though, are predictable, and most of them cost money before they show any benefit. Here are eight checks worth running before you commit to a tool or a developer.

1. Decide what a price change should actually trigger

Start at the end. If a competitor drops a price by four percent, what happens in your business? If the honest answer is nothing, because your margins are fixed by an importer and you cannot move, then daily monitoring is a report nobody reads.

Useful triggers tend to be narrow: a competitor going out of stock on an item you hold, a price crossing below your cost, or a product appearing that you do not carry. Write those three rules down first. They determine everything else, including how often you need to collect and how much you should spend.

2. Count your real SKUs, not your catalogue

A catalogue of five thousand items usually contains a few hundred that generate most of the revenue and most of the competitive pressure. Monitoring those hundreds daily is cheap. Monitoring everything daily is roughly ten times the cost for information you will not act on.

Pull last year’s sales by item, take everything above the median, and start there. You can always widen the list once the system proves useful.

3. Check whether the data is even collectable

Not every competitor publishes prices. In several categories, particularly building materials, industrial supply and wholesale, the price sits behind a login or a quote form. No tool will get that for you legitimately, and any vendor promising it is describing something you do not want to be part of.

Spend an hour checking your top ten competitors by hand before you buy anything. If seven of them hide prices, the project is a market research exercise, not an automation one.

4. Understand why the same page shows different prices

This is the check most projects skip and then spend a month debugging. Ecommerce platforms serve different content depending on where the request appears to come from. Currency, delivery estimates, tax display and sometimes the price itself change with the visitor’s location. A server collecting from abroad sees a version of the page that no local customer ever sees.

The consequence is a database full of numbers that are technically accurate and commercially useless. Fixing it means collecting from an address that looks like a normal local connection rather than a hosting company. For a catalogue of a few hundred items, an affordable isp proxy plan does this for less than the cost of one staff hour a month, and it gives you a fixed address, which matters more than it sounds. A fixed address behaves consistently, can be added to allow lists, and does not trip the verification screens that rotating connections run into.

Providers such as ProxyWing sell these as ISP or static residential addresses. The naming is confusing across the market, but the property you want is simple: registered to a consumer internet provider, hosted on stable infrastructure, and located in the country your customers are in.

5. Match products properly before you compare anything

Two listings with the same brand and model number are frequently not the same product. Different bundle contents, different warranty terms, grey imports and different pack sizes all produce price gaps that mean nothing.

Match on a manufacturer code where one exists, and treat title matching as a suggestion that a human confirms once. A wrong match is worse than no match, because it produces a confident alert about a competitor undercutting you when they are selling a different thing.

6. Plan where the data lands

Collected prices are only useful next to your own numbers. If your inventory and purchase costs live in an ERP or a POS system, the monitoring output should end up in the same place, keyed by the same item codes.

A separate dashboard nobody opens is the usual failure mode. The version that works is boring: a daily table inside the system your buyer already uses, with three columns, showing only items where a rule fired.

Where the data goes	Who looks at it	Does it get used?
Standalone dashboard	Nobody after week two	Rarely
Daily email of all prices	Skimmed, then filtered to a folder	Rarely
Alerts only when a rule fires	Purchasing	Usually
Column inside the ERP item screen	Whoever sets prices	Consistently
Weekly summary for management	Owner or manager	For decisions, not for daily work

7. Budget for maintenance, not just setup

Competitor sites change layout a few times a year, and every change breaks whatever is reading them. Budget a few hours a month for repairs, or choose a vendor whose price includes that work and check what their support response time actually is.

The failure mode to watch for is silent breakage. A collector that returns zero results looks the same as a competitor with no price changes. Add a simple check that alerts you when collection volume drops sharply, and you will catch it the same day rather than the next quarter.

8. Know the rules you are operating under

Reading publicly published prices is normal commercial practice. A few boundaries still apply, and they are worth knowing before you start rather than after a complaint.

• Do not create accounts to reach prices behind a login. That moves you from reading public information to breaching terms you agreed to.
• Collect at a modest rate. A few requests per minute per site is invisible and considerate. Hundreds per second is a different conversation.
• Do not collect personal data. Reviews with names, seller contact details and customer questions are not part of a price monitoring project and bring obligations you do not want.
• Do not republish competitor content. Using a price to inform your own decision is different from displaying their catalogue on your site.

Who should own this inside the business

Price monitoring fails more often for organisational reasons than technical ones. It gets built by whoever is most comfortable with software, which is usually not the person who sets prices, and then it drifts because nobody depends on it.

The arrangement that survives is simple. The person responsible for purchasing or pricing owns the rules and the competitor list, and reviews both once a quarter. Whoever maintains your systems owns the collection and the alerting. If a single person holds both roles, which is common in a smaller company, write the rules down anyway, because the knowledge otherwise leaves with them.

One more thing worth agreeing early: what happens when a competitor is cheaper. Businesses that have not decided this in advance tend to react item by item and end up in a price war they did not intend to start. A rule such as matching only on the top fifty items, and only within a defined margin floor, keeps the system useful without letting it drive the business.

What a sensible first month looks like

Week one, pick the top two hundred items and the five competitors that matter. Week two, get collection running once a day and check the numbers by hand against the live sites. Week three, write the two or three alert rules you defined at the start. Week four, put the output where the person who sets prices already works.

That is a system that pays for itself on a single missed stockout. Everything more elaborate can wait until this version has been running for a quarter and someone asks for more.

FAQ

How much does this cost to run for a small business?

The collection layer is usually the cheapest part. A modest address plan and a small server cover a few hundred items comfortably. The real cost is the few hours a month of maintenance.

Can we use a free proxy or a VPN instead?

For a one off check, a VPN is fine. For scheduled collection, free proxies are unreliable and often unsafe, and consumer VPN endpoints are widely recognised, so results become inconsistent.

Should we monitor marketplaces or competitor websites?

Marketplaces first, because that is where price comparison actually happens for most buyers. Add direct competitor sites afterwards for the items where you compete on service rather than price.

How often should prices be collected?

Daily is enough in most categories. Electronics and anything tied to currency movement can justify twice a day. Hourly monitoring is rarely worth what it costs.

What if a competitor blocks us?

Slow down first, because the usual cause is request volume. If a site consistently refuses automated access, respect that and reallocate the effort to sources that do not.