Arunesh Mathur

Who they are

The 2019 paper lists Mathur at Princeton University. His co-authors include Gunes Acar, who by 2023 was at Radboud University and co-supervised the countdown-timer experiment described on the page for Jelmer Tiemessen. This page records nothing about Mathur beyond the cited work.

The crawl visited the product pages of 11,286 shopping sites, about 53,000 pages in all, and found 1,818 dark pattern instances on 1,254 of the sites, or about 11 per cent. It also identified 22 third-party entities that supply dark patterns to shops as a service, and proposed a taxonomy of what makes a pattern manipulative.

Under the heading of urgency, it separated countdown timers, which show a deadline moving toward zero, from limited-time messages, which claim a sale will end soon without giving a deadline. The team found 393 countdown timers. To test them for deception they revisited sites over time, stitched the screenshots into videos and watched what the clocks did. A timer was counted as deceptive if it reset after reaching zero with the offer still valid, or if the offer it claimed was expiring was still valid after the timer expired. By that test, 157 timers on 140 sites were deceptive.

What counting at scale adds

A lab experiment can show what a false timer does to a handful of people. It cannot show how many shops run one. The 2019 crawl answers the second question. It treated the countdown timer as one pattern among many in a taxonomy of what makes a design manipulative, and it tested for deception by behaviour over time and not by a timer's look. A clock that reset with the offer still on, or whose offer outlasted it, was counted as deceptive. That made the judgement checkable by anyone who revisits the page.

The figures need careful reading. Of 393 countdown timers, 157 were deceptive, on 140 sites. Those counts come from a crawl of the product pages of 11,286 shopping sites, so they describe what that crawl found on those pages in 2019, and not every shop, and not today. A smaller hand-built sample of 60 timers in the Radboud thesis found 27 judged deceptive, 22 of them with deadlines under 24 hours. The two counts use different methods and measure different things, so they cannot be laid side by side.

How the paper has been used

This publication cites the paper's counts for how common false urgency was on shopping sites. A 2023 screening of 399 online shops by the European Commission and consumer authorities from 23 member states, Norway and Iceland, found 42 using fake countdown timers. That is a separate exercise with its own method, and neither asked sellers why they ran the timers.

The paper also keeps an honest limit on its own claim. It notes that opinions on countdown timers vary among experts and users, and it leaves open how far users perceive them as manipulative. For a reader asking whether timers sell more, the crawl supplies prevalence and no measure of effect.

Key papers

Mathur, A., Acar, G., Friedman, M. J., Lucherini, E., Mayer, J., Chetty, M., & Narayanan, A. (2019). "Dark Patterns at Scale: Findings from a Crawl of 11K Shopping Websites." Proceedings of the ACM on Human-Computer Interaction, 3(CSCW), Article 81. doi:10.1145/3359183.

Key ideas

Dark patterns can be measured at scale by crawling, not only studied one at a time in a lab. The paper notes its own limit on the countdown timers: opinions about them vary among experts and users, and it leaves open how far users perceive them as manipulative. See Dark Patterns and Deceptive Countdown Timers.

---