Well it's O( f(n) * g(n) ) or O( max( f(n), g(n) ) )
There can be three cases possible with f(n) and g(n) -
Case A : f(n) > g(n)
In this case we take O(f(n)) the complexity of the algorithm as g(n) is a lower order term, we drop it.
Case B : f(n) < g(n)
In this case, we take O(g(n)) the complexity of the algorithm as f(n) is a lower order term, we drop it.
Case C : f(n) == g(n)
Time Complexity can be either O(g(n)) or O(f(n)) (which is equal asymptotically).
Hence, max(f(n), g(n))