Hmm no pretty normal I think.
That's why I say there's some debate over this - you HAVE to go to silly lengths to ensure the bowl/depression is deep enough otherwise they will eventually fall off - most people don't go that far, and thus you end up with the school of thought that "it's fine on the sand" because it's just easier than rearranging your whole aquascape. And yeah, it's sort of fine .. but it sort of isn't, either. It's better in that they can't fall off and get injured, but if you ever see tridacnids in the wild - with the exception of maybe gigas, squamosa and derasa, who are sand dwellers - they are always embedded in rock. IMO this is where they belong, because that byssal opening is pretty big and it's an area that pests and predators can get into if it's not attached to rock. It's not a guarantee that there will be problems if not on the rock but it IMO it does help a little.
Part of the problem is they shoot water out with such force that it moves them. So if the attachment point isn't 100% secure, they'll move. I don't think they really intend to move this way or that, I think it's more of a random result of their water "surges".
Another thing to keep in mind if on the sand is that they might still attach to the first hard thing they find - the bottom glass - and it can be very difficult to remove them in that case. You can remove clams after they are attached but it takes a lot of practise and it's a risky move, there is potential for injuring them this way.
I realize I'm firmly in the minority in this recommendation, but, it IS my recommendation and I'm sticking with it.
Since adopting this approach myself I've had way better clam longevity than I ever did when I just let them sit in the sand. My oldest clam is a crocea around 7 years old right now.