What a tracker actually records
How price tracking works is simpler than most people assume, and the limits are more interesting than the mechanism.
A tracker visits a listing at intervals, records the price it displayed, and stores the result against a timestamp. Repeat that for months and you have a history the listing itself does not show: the range the price has moved in, the level it usually returns to, and where today falls against both.
That is the whole capability, and it is enough to answer the one question a product page structurally cannot: what does this thing normally cost? Which is the question the pillar this sits under is about, is this a good deal.
Where the data comes from
Three sources, with different strengths.
Scraping the public listing. The most common. A tracker requests the page as a visitor would and reads the price from it. Broad coverage, and vulnerable to page changes, rate limits and regional variation.
Retailer APIs and affiliate feeds. More reliable and structured where they exist, and usually narrower in what they expose. Some carry only a subset of a catalogue, and some update less often than the site does.
Aggregated user reports. Rare on its own, occasionally used to fill gaps.
Sampling frequency varies and matters less than people expect. A tracker sampling several times a day will catch most movements; one sampling daily will miss a short flash sale and still describe the band correctly. The band is what the history is for, and short spikes are noise against it.
What matters far more than frequency is which listing is being sampled, which is the subject two sections down.
What tracking cannot see
Being precise here is what makes the rest trustworthy.
What anyone paid. A listing can display a price for six months and sell nothing at it. The chart records the shelf price, not the transactions, and no public tracker can tell the difference. This is the honest limit of all price history and the reason a tracked average is a strong signal rather than proof.
Prices behind a condition. Member-only pricing, coupons that must be clicked, codes entered at checkout, business accounts, and anything that only appears in a basket. A tracked price is the shelf price rather than the till price, and on some listings the gap is substantial.
Regional and currency variation. The same product carries different identifiers, prices and histories in different countries.
Stock. A gap in a line usually means the product was unavailable. An out-of-stock product has no price rather than a low one, and a line drawn straight across a gap can make a level look like it held far longer than it did.
The future. A chart showing a drop every November is evidence about past Novembers. Treating it as a forecast is the most common misuse of an otherwise honest tool.
The identity problem, which is the real one
The most common way a price history misleads has nothing to do with sampling. It is that the history is not about your item.
One product page can carry several sizes, colours, capacities, pack counts and conditions, each with its own price. A tracked line that mixes a single unit with a four-pack, or new stock with refurbished, produces a shape that describes no purchase anybody could make: the lows are the cheap variant and the highs are the expensive one.
The same applies across a product's life. A listing revised to sell a newer model keeps its history, so a chart can span two different products with a step in the middle that looks like a price change.
Good tracking pins the exact configuration first and then measures it, and says so when it cannot. A tracker that will not tell you which variation it followed deserves more caution than anything the line is doing, and a sudden step with no product news behind it is frequently a variation change rather than a price movement. How that grouping works, and how to check it on a listing, is covered in Amazon variation mixing.
What a good tracker does that a bad one does not
Four differences, and none of them is sampling frequency.
It states which variation it followed. The single most important property, and the one most often missing. A history that silently spans several configurations is worse than no history, because it looks authoritative.
It marks out-of-stock periods rather than interpolating across them. A gap drawn as a flat line turns an absence of price into an apparent stable one.
It distinguishes the seller. On a marketplace listing the price depends on who holds the buy box, so a line that mixes sellers is measuring several offers as though they were one.
It says what it cannot see. A tracker that presents a shelf price as the price, with no note about coupons and conditional discounts, is overclaiming in a way that matters most at exactly the moment someone is deciding.
Judging a tracker on those four is more useful than judging it on how often it samples, because all four are about whether the line means what it appears to.
Using it well
Four habits turn a chart from interesting into decisive.
Open the year view first. The seasonal pattern only exists at that scale, and the range over twelve months tells you immediately whether this price moves enough to be worth timing at all.
Find the band, not the average. Where the line spends most of its time is the product's real price. The average is pulled by spikes in both directions and describes a level the product may rarely have held.
Read the shape of the recent past. A step down that held for months is a price cut. A dip that snapped back within days was a promotion. They look identical if you only see today. The detailed version of this is in how to read a price history chart.
Set alerts from the history rather than from hope. A threshold taken from the bottom of the observed band will fire. One taken from what you wish the thing cost will not, and the decision stays open indefinitely.
To pull the tracked line for one specific listing rather than reading about the method, use Amazon price history.
One final note on expectations. Price tracking is a measurement tool rather than a prediction tool, and almost every disappointment with it comes from asking it the second question. It can tell you with confidence what a product has cost and what today looks like against that. It cannot tell you whether next month will be cheaper, and a tracker presenting a forecast is selling something the data does not support.
Used for what it does, it is the single most useful instrument in online shopping: it converts a price from a number into a position, and a position is something you can act on.
A practical closing note on alerts. Two thresholds are worth setting on anything you are waiting for: one at the bottom of the observed band, which is the realistic buy signal, and one slightly above it, which gives you warning that the price is moving in the right direction. A single alert at a number the product has never reached is the most common way people wait indefinitely for a purchase they intended to make.
