Invalid traffic & fraud · also called Behavioural biometrics, interaction analysis
Behavioral analysis
Behavioral analysis in fraud detection is examining how a visitor interacts, such as mouse movements, taps, scrolling, typing rhythm and session timing, to tell real humans from bots and scripted activity.
Humans are messy. They move a mouse in curves with small hesitations, scroll unevenly, type in bursts, pause to read and wander between pages. Bots, even good ones, tend to be too smooth, too fast, too regular or oddly repetitive. Behavioral analysis measures these patterns.
Signals include mouse trajectories and acceleration, touch pressure and swipe dynamics on phones, device motion from sensors, keystroke timing, time to first interaction, click position relative to buttons, scroll depth and rhythm, and session-level behaviour such as page sequences and time of day. At the population level, analysts look for unnatural uniformity: thousands of sessions with the same duration, identical click coordinates or engagement that spikes on the hour.
It is a central tool for SIVT, used alongside device fingerprinting, IP reputation and supply-chain checks. Sophisticated bots now replay recorded human movements or add randomness, and human fraud farms involves real humans, so behavioural analysis works best combined with other evidence. It must also be implemented with care for privacy and accessibility, since some genuine users (for example, those using assistive technology) behave differently.
Think of it like this
It is like a shop assistant who can tell a browsing customer from a shoplifter by how they move, not by what they look like.
An example
A lead form receives 5,000 submissions a day; behavioral analysis shows 1,800 were filled in under 3 seconds with no mouse movement and identical keystroke intervals, so they are rejected as bot-submitted.
Related terms
Device fingerprinting
Device fingerprinting is identifying or characterising a device from the combination of its technical attributes, such as browser, fonts, screen, graphics and network traits, instead of a stored cookie or ID.
Bot traffic
Bot traffic is visits, ad impressions or clicks generated by automated software rather than people, ranging from honest search crawlers to fraudulent bots built to look human.
Fraud detection
Fraud detection in advertising is the process and technology of identifying invalid or fraudulent impressions, clicks, installs and conversions, so they can be blocked in advance or excluded from billing.
Headless browser
A headless browser is a web browser that runs without a visible window, controlled by code; widely used for testing and scraping, and by bots to fake human visits and ad views.
Lead generation fraud
Lead generation fraud is submitting fake, stolen or incentivized form fills, such as sign-ups, quote requests or demo bookings, so that advertisers pay for leads that are not genuine prospects.
Sources: MRC Invalid Traffic Detection and Filtration Guidelines Addendum (2020 update)