Well, there is a difference between tsunami from earthquake and tsunami from a huge landslide just across the bay.
I'd say proximity and directionality.
An earthquake/tectonic event (excepting in the events of the '2012' film, sheesh) makes a section of seabed shift up/down/round-and-round/whatever, and may spark underwater landslides, but even in coastal waters is going to (
in general) cause a relatively small height (say 10s of metres, as recently seen) of total water displacement, across a large area, and what's more that energy gets shoved out in all directions (not equally so, but generally). Meaning that if it's quite close to a shore, that shore will get almost that amount of wave, but it'll lessen (excepting where it 'refracts' around underwater and shoreline features, constructively building) the further away the wave goes. Unless it's an extremely in-shore event (in which case the raising/lowering of the landscape will raise/drop the coastal area as well), there'll be dissipation.
If you have, on the other hand, a deep lake with an tall, unstable hillside on one edge, then a landslip (caused by anything, but possibly also an earthquake) might well dump an amount of soil/rock/etc into the lake... This could represent a significant proportion of the volume of the lake, dumped in at one narrow band at one edge, and you end up with have half a lake jumping out of its bath, Archimedes-like, onto the opposing shore and doing a watery 'Euraka!' dance all over the land on the other side. There's no actual room for any meaningful dissipation and it's almost all aimed in one direction. Deep and narrow lake + co-operative geography + huge landslide => 500m wave.
That's skimming over some of the issues, but as a summary it'll probably do.