The electromagnetic ‘ponderomotive force’ was discovered in 1957 in a radio-frequency experiment by Boot and Harvie 1. The force arises due to gradients in an inhomogeneous oscillating electric field, and can be used to control and accelerate electrons, with applications in ion traps, plasma accelerators and high-harmonic generation. However, being a second-order nonlinear effect in the applied field, the ponderomotive force is relatively weak. The availability of high-power, short laser pulses has enabled new practical realizations of this effect 2. In a Letter in this issue, Batelaan and co-workers predict that metal nanoantennas can produce sufficient ponderomotive force to deflect electrons 4. The phenomenon relies on the surface plasmon resonance properties of the nanoantenna, which produces a high local field enhancement combined with large field gradients in a small volume around the nanoantenna. To achieve the required field strengths, they propose to use high energy laser pulses of femtosecond time duration, tuned to the surface plasmon resonance wavelength. These short laser pulses offer yet another advantage, as they allow producing an ultrafast modulation of an incident electron beam. Such an electron switch would be of interest for a range of applications, for example ultrafast electron microscopy 4. Ultrafast electron microscopy is a powerful tool for characterizing chemical and physical processes at nanometer length scales and ultrafast time scales. Currently, the generation of ultrafast electron pulses is done by directly exciting the electron emission source with a high-power UV laser pulse. Plasmonic electron switches may be a feasibly alternative route toward the development of ultrafast electron beams at higher repetition frequencies and using lower-power femtosecond oscillators. As a first step, the authors measured the coherence time of surface plasmon modes in an array of nanoparticles, so-called optical antennas 3. A short, resonant laser pulse can drive a coherent plasmon oscillation, however this oscillation rapidly dephases via collisions of the electrons (intraband) or through excitation of electron-hole pairs (interband). Surface plasmons in isolated metal nanospheres have a dephasing time of several femtoseconds, depending on the type of metal antenna under study and details of the damping mechanism. Detailed measurements of plasmon dephasing were pioneered in the 1990s by Franz Aussenegg and co-workers, who used high harmonic generation as a nonlinear process to make an autocorrelation of the plasmon pulse 5. High-harmonic generation in nanoparticles is complicated, since second-harmonic requires noncentrosymmetric structures, while third harmonic signals are comparably weak. Becker et al. use a different, more direct method of measuring both the amplitude and phase of the plasmon field using cross-correlation of the plasmon field with a reference pulse with known width taken directly from the laser. The prospect of electron beam modulation using ultrafast plasmon fields is promising as a fundamental phenomenon and may further open up the field of ultrafast electron microscopy. Both the coherence time and local field enhancement can be tailored by the geometry of the nanoantenna, allowing for a rational design of nanostructures optimized for specific applications 6. Of particular interest will also be the application of techniques from coherent control to achieve local plasmonic hot-spots in both space and time 7, potentially resulting in even stronger field gradients at shorter time scales. Several hurdles will have to be overcome, related to the high field strengths required and the breakdown of metal nanoantenna arrays under such high power excitation. Similar problems are encountered in higher-harmonic generation experiments, and require careful optimization of material properties of the antennas and substrate. However, if successful, the plasmonic antenna approach may result in a new generation of electron modulation devices.