Coastal cliff erosion is caused by a combination of marine forcing and sub‐aerial processes, but linking cliff erosion to the environmental drivers remains challenging. One key component of these drivers is energy transfer from wave–cliff interaction. The aim of this study is to directly observe cliff ground motion in response to wave impacts at an individual wave scale. Measurements are described from two coastal cliff sites: a 45‐minute pilot study in southern California, USA and a 30‐day deployment in Taranaki, New Zealand. Seismometers, pressure sensors and video are used to compare cliff‐top ground motions with water depth, significant wave height (Hs) and wave impact types to examine cliff ground motion response. Analyses of the dataset demonstrate that individual impact events can be discriminated as discrete events in the seismic signal. Hourly mean ground motion increases with incident Hs, but the largest hourly peak ground motions occurred across a broad range of incident Hs (0.9–3.7 m), including during relatively calm conditions. Mean hourly metrics therefore smooth the short‐term dynamics of wave–cliff interaction; hence, to fully assess wave impact energy transfer to cliffs, it is important also to consider peak ground motion. Video analyses showed that the dominant control on peak ground motion magnitude was wave impact type rather than incident Hs. Wave–cliff impacts where breaking occurs directly onto the cliff face consistently produced greater ground motion compared to broken or unbroken wave impacts: breaking, broken and unbroken impacts averaged peak ground motion of 287, 59 and 38 μm s−1, respectively. The results illustrate a novel link between wave impact forcing and cliff ground motion response using individual wave field measurements, and highlight the influence of wave impact type on peak energy transfer to coastal cliffs. © 2019 John Wiley & Sons, Ltd. © 2019 John Wiley & Sons, Ltd.