We determine the pre-asymptotic critical behavior at the quantum ferromagnetic transition in strongly disordered metals. We find that it is given by effective power laws, in contrast to the previously analyzed asymptotic critical behavior, which is valid only in an unobservably small region. The consequences for analyzing experiments are discussed, in particular ways to distinguish between critical behavior and Griffiths-phase effects.PACS numbers: 64.70. Tg, 05.30.Rt, 75.40.Cx, 75.40.Gb The ferromagnetic quantum phase transition in metals has received a lot of interest in recent years. In clean systems it is now well established that the transition is generically first order. This has been observed experimentally for ferromagnets as diverse as the d-electron metals MnSi and ZrZn 2 [1, 2], and various uranium-based compounds [3]. It is in sharp contrast to early theories that predicted a second-order transition with mean-field exponents for both clean and disordered quantum ferromagnets [4], but in excellent agreement with a theory that takes into account the coupling between the magnetization and soft or gapless fermionic excitations in metals [5,6]. Disordered systems, which include almost all magnets where the quantum phase transition is triggered by chemical composition, are not as well understood. Experimentally, continuous or second-order transitions have been observed in a variety of systems, but attempts to determine exponents have yielded very different results for different systems, and even for different analyses of the same system [7][8][9][10]. Theoretically, the same framework that predicts a first-order transition in clean systems predicts a second-order one in the presence of quenched disorder [11][12][13]. Again, the coupling of gapless excitations (in this case, diffusive particle-hole excitations) leads to strong deviations from the prediction of Hertz theory [14]. The asymptotic critical behavior is given by non-mean-field power laws modified by multiplicative log-normal terms. Comparisons between this asymptotic critical behavior and experimental results are overall not satisfactory. Another important development has been the study of quantum Griffiths effects that are expected to be present in systems with quenched disorder [15,16]. These are due to rare regions devoid of disorder, occur in a whole region in the phase diagram, and are characterized by non-universal power laws. The interplay between Griffiths-phase effects and critical phenomena is incompletely understood and makes the interpretation of experiments difficult, see the discussion at the end.The quantum critical point is difficult to determine experimentally, and the distance from it can be hard to control. As a result, the experimental situation is much more difficult than in the case of classical phase transitions. For the latter, it is known that in almost all cases an observation of the asymptotic critical behavior requires measurements over two decades or more in a region within less (in some cases substantially less) t...