Dwell Time | ScrapeRouter

Dwell time usually means how long someone stays on a page after clicking a search result before going back. In scraping, people also use it more loosely to mean how long a browser session sits on a page before the next action, which matters because real users don’t hit pages with perfect machine timing.

Examples

A basic example in browser automation is adding a small, realistic pause after navigation before clicking the next element.

import random
import time
from playwright.sync_api import sync_playwright

with sync_playwright() as p:
    browser = p.chromium.launch()
    page = browser.new_page()
    page.goto("https://example.com")

    dwell_time = random.uniform(2.0, 5.5)
    time.sleep(dwell_time)

    print(f"Stayed on page for {dwell_time:.2f} seconds")
    browser.close()

If you're calling a scraping API with browser actions, dwell time is often the wait between steps, not just a blind fixed sleep.

{
  "url": "https://example.com/product/123",
  "browser": true,
  "actions": [
    {"type": "wait_for", "selector": "h1"},
    {"type": "wait", "ms": 3200},
    {"type": "click", "selector": "button.show-more"}
  ]
}

Practical tips

Don’t treat dwell time as a magic anti-bot trick: fixed 3-second sleeps everywhere just make jobs slower and still look fake.
Prefer event-based waiting over arbitrary waiting: wait for a selector, API response, DOM change, or network idle when possible.
Add variance when you do need pauses: real browsing patterns are uneven, not perfectly spaced.
Keep the page type in mind: search results, product pages, and infinite scroll feeds all have different natural timings.
Measure the cost: extra dwell time across thousands of pages turns into real runtime and proxy spend fast.
If a target is sensitive, combine dwell time with other things that matter more: good fingerprinting, sane concurrency, session continuity, and IP quality.
In production, the question is usually not "can we wait longer" but "what is the minimum wait that keeps extraction stable."

Use cases

Search and SEO analysis: estimating whether users actually engaged with a result or bounced right back.
Browser automation for scraping: spacing interactions so a workflow behaves more like an actual session.
Dynamic page extraction: giving client-side content, lazy-loaded elements, or anti-bot challenges time to settle before extraction.
Session simulation: keeping a login or browsing flow from looking absurdly fast.
Router-based scraping setups: if you use ScrapeRouter with browser rendering, dwell time is one of the small control knobs you may use, but it works best alongside retries, provider routing, and proper wait conditions.