That question is kind of tricky, mainly because there are many assumptions here. So, let’s try to reframe that by making some precisions about your laser beam itself and the conditions in which measurements are performed.
First, we assume here that your beam is reaching the detector surface at normal incidence. So, the cross-section of the beam and the detector surface are both in one single plane. Second, we consider that your beam is perfectly centered in the aperture of the power meter. Third, the power distribution of your beam can vary largely according to its shape and profile. So, let’s assume here that your beam has one of these profiles: Gaussian or flat-top, which are the most common types.
For the flat top beam, the power is theoretically evenly distributed along its cross-section and outside of this section, the power drops to zero. The diameter of the beam is clearly defined. On the other hand, the Gaussian beam mimics the shape of a three-dimensional Gaussian function which has a gradual decrease from high intensity in the center to gradually lower intensity. Therefore, to define a Gaussian beam’s diameter we must rely on some conventions since a Gaussian function does not have a clear edge from which you can measure the beam width. Many definitions exist, but the diameter is commonly defined as the width of the beam at the point where the intensity drops to 1/e² (about 13.5%) of its maximum value. This point is known as the beam waist radius. Accordingly, the diameter is twice that.