Most teams pick the wrong thing to build first — not because they lack ideas, but because nobody wrote down what the bottleneck actually costs. Here is the test.