# Scraper bots vs scalper bots: how to tell them apart in traffic logs | Bot Traffic Index

Source: https://bottrafficindex.com/blog/2026-10-04-scraper-vs-scalper-bots.html

Oct 4, 2026 - Bot Traffic Index


Both hit product pages hard, but they want different things. Scrapers read broadly and evenly; scalpers strike narrowly and fast. The request pattern gives each away.

## Two bots, two business models

In the traffic logs, scraper bots and scalper bots can look similar: bursts of requests to product pages, odd user agents, no purchases. But they want different things, and the difference matters because the response is different. Scrapers want your data: prices, inventory, product details, images. Scalpers want your inventory itself: limited products bought faster than humans can click. Misidentify one as the other and the countermeasure misses.

## The scraper pattern: broad and even

Scrapers read catalogs. Their signature is breadth: many distinct product pages, often in sitemap order or sequential IDs, at a steady pace. A scraper working through 5,000 SKUs will request them methodically, sometimes respecting crawl delays, sometimes not. Session durations are long, page depth is shallow (product page in, product page out, rarely the cart), and the same IP or IP range returns day after day to re-read the catalog.

Scrapers also read the supporting pages: category listings, search results with pagination, and sometimes the cart API to check live inventory without rendering the page. Image requests are a giveaway; humans load images as part of the page, scrapers often request image URLs in bulk separately. If a session downloads every product image but never views a collection page, it is not shopping.

## The scalper pattern: narrow and fast

Scalpers do the opposite. They care about a handful of high-demand products and they hit those pages with extreme focus: the same product URL requested hundreds of times in minutes, inventory-check endpoints polled aggressively, add-to-cart attempted the instant stock appears. Where the scraper is a librarian, the scalper is a sprinter.

The timing signature is the clearest tell. Scalper traffic spikes around drop times and restocks, goes quiet between them, and shows inhuman speed in the critical window: page load to add-to-cart in under a second, checkout form completion faster than typing allows. Multiple sessions from related IPs or with rotating fingerprints all targeting the same SKU at the same moment is a coordinated operation, not coincidence.

## Where they overlap, and how to split them

Both may spoof user agents, rotate IPs, and run headless browsers. The split comes from intent signals in the request sequence. Ask three questions of any suspicious session: How many distinct products did it touch (many suggests scraper, one suggests scalper)? Did it ever attempt cart or checkout actions (scalpers do, scrapers rarely do)? What is the request cadence relative to inventory events (spikes at restocks suggest scalper)?

Inventory-check endpoints are the richest signal on most stores. Scrapers poll them across the catalog on a schedule. Scalpers poll them for specific SKUs at high frequency around drops. Log these endpoints separately from page views and the two populations separate themselves.

## Different bots, different responses

Scrapers are usually handled with access controls: rate limits, bot management rules, terms-of-service enforcement, and sometimes deliberate data poisoning for the worst offenders. The goal is to raise the cost of reading without hurting real shoppers, who never request 5,000 product pages in an hour.

Scalpers need purchase-path defenses: purchase limits per customer, queue systems for drops, checkout velocity checks, and payment-level signals. Blocking a scalper's IP just moves them to the next proxy; the durable defenses verify the buyer, not the browser. CAPTCHAs at the add-to-cart moment for hyped products remain unglamorous and effective.

## Log for the distinction

To tell them apart routinely, the logs need: per-session distinct product counts, cart and checkout attempt flags, inventory-endpoint request rates per SKU, and timing from page view to cart action. With those four fields, the scraper-vs-scalper question answers itself for most sessions, and the right countermeasure follows from the answer.

## Frequently asked questions

### What is the difference between a scraper and a scalper bot?

Intent and target. Scrapers harvest information, prices, catalogue data, and content, at reading speed. Scalper bots transact: they create accounts, hold carts, and complete purchases of limited stock in seconds. The defences, and the harm, are different for each.

### How do they look different in logs?

Scrapers show systematic page coverage with no buying behaviour. Scalpers cluster around drops and product launches, with account creation spikes, instant cart to checkout times, and success concentrated on the scarcest items. A traffic pattern that buys is not a scraping pattern, however automated it is.

### Why does the distinction matter operationally?

Because the teams and tools differ: scraping is a data protection and infrastructure question, scalping is an inventory fairness and fraud question. Blocking a scraper's method does nothing to a scalper queueing for a drop, and queue defences do nothing about catalogue harvesting.

### Can one operation run both?

Yes, commonly. Scraping maps the catalogue and prices first, and the same operator automates purchases of the items worth reselling. Finding one pattern is a good reason to look for the other in the same traffic, with the same time window.