Doherty Threshold¶

Status: emerging
Last updated: 2026-06-02
Sources: Lawsofux.Pdf
Tags: [ux-design, design-principles, heuristics, doherty-threshold, response-time, performance, feedback]

Summary¶

The Doherty threshold holds that productivity rises sharply when a computer and its user interact at a pace below about 400 milliseconds, so that neither waits on the other (Yablonski, 2024). Applied to UX, it sets a target for system responsiveness: feedback should be near-instant, and where real processing takes longer, the interface should manage perceived performance with progress indicators and immediate acknowledgement. Yablonski traces the threshold to a 1982 IBM study by Doherty and Thadani.

Body¶

Context¶

Yablonski (2024), in the Doherty Threshold chapter of Laws of UX, applies research on computer response time and productivity to interaction design. The chapter states the threshold, gives its origin, and derives guidance for responsiveness and perceived performance. Within this knowledge base the article connects to Usability And User Experience (efficiency and responsiveness) and the attention/feedback concerns of Information Processing; it is one of the interaction laws in Laws Of Ux.

Key Points¶

Yablonski states the threshold as: productivity soars when a computer and its users interact at a pace (under 400 ms) that ensures that neither has to wait on the other (PDF pp. 139–150, orig. pp. 119–130). He traces it to a 1982 study by IBM employees Walter J. Doherty and Ahrvind J. Thadani, "The Economic Value of Rapid Response Time" in the IBM Systems Journal, which challenged the then-standard two-second response target and found that productivity increases more than in direct proportion to a decrease in response time once the threshold falls under 400 ms (PDF pp. 139–150, orig. pp. 119–130).

The design implications concern both actual and perceived speed. Systems should respond within the threshold wherever possible, and where an operation genuinely takes longer, the interface should preserve the sense of responsiveness — acknowledging input immediately, showing progress indicators, and using techniques that shorten perceived wait time (the chapter cites progress animations and optimistic UI of the kind used to mask latency) (PDF pp. 139–150, orig. pp. 119–130). The chapter relates the experience of fast, fluid interaction to a state of flow.

Conclusion¶

Yablonski (2024) concludes that response time is an experience variable, not just an engineering metric: keeping interaction under roughly 400 ms keeps users engaged and productive, and when true speed is not achievable, managing perceived performance preserves much of the benefit.

References¶

Yablonski, J. (2024) Laws of UX: Using Psychology to Design Better Products & Services. 2nd edn. Sebastopol, CA: O'Reilly Media. yablonski2024lawsux

Doherty, W.J. & Thadani, A.J. (1982) 'The economic value of rapid response time', IBM Systems Journal. To be validated.

Open Questions¶

How does the 400 ms threshold translate to modern interaction types (streaming, AI generation, network-bound actions) where sub-threshold response is often impossible?
How much of the productivity benefit can be recovered through perceived-performance techniques when actual response time exceeds the threshold?