Two failure modes come up constantly with streamer fishing, and both are easy to diagnose once you know what to look for.
The fly is too high. The most common signal is a follow that ends early. A trout chases the fly and then turns off without eating. Sometimes you can see this in clear water, and sometimes you just feel the fish bump or pull and then nothing. What often happens is that the fly swings upward through the retrieve and the fish loses interest when it rises out of the comfortable feeding zone. Other clues: you are not getting any contact with the bottom in a section you know has deep holding water, or your fly feels weightless through the whole retrieve with no sense of the current pushing on it from below.
Fix it by switching to a sinking-tip or full-sinking line, shortening the leader, counting down longer before you start stripping, slowing the retrieve, or adding an upstream mend after the cast to delay line tension. You can also try a heavier fly if you are already on a sinking-tip.
The fly is too deep. This one is obvious and expensive: the fly snags bottom. You strip and the line stops abruptly, then comes back without the fly or with a piece of moss on the hook. If it happens repeatedly in the same run, you are running the fly too low.
Fix it by switching to a floating or intermediate line, lengthening the leader (which allows the fly to ride higher), retrieving faster to create more upward tension, or changing the cast angle to across or downstream to reduce sink time. Switching to a lighter or unweighted pattern also helps if the fly itself is contributing to the depth problem.
The diagnostic tool that helps most with both failure modes is watching your flies in shallow, clear water before you fish the deep run. Drop a weighted Beadhead Woolly Bugger into ankle-deep water on a short line and watch what happens when you strip it, let it swing, and pause the retrieve. You will see exactly how quickly it drops, how fast it rises on the strip, and how it behaves on a pause. That visual information translates directly to your mental model of what the fly is doing in the water you cannot see.
One last note on follows: a fish that chases the fly all the way to the bank and turns off at the last second is giving you information. That fish was interested enough to move but did not commit. More often than not, the fly path through the water was slightly off – either the depth, the speed, or the angle did not quite match what the fish wanted to do. Follows are feedback, not failures.