If you think it may be a tautology, then it would mean that it is impossible for the antecedent to be true and the consequent false, so proof by contradiction could be a way to go. The one given below is quite informal.
Suppose for contradiction the negation of the conditional, i.e. suppose that $(\exists x)(\forall y)P(x,y)$ is true and the negated consequent, i.e. $ \neg (\forall y)(\exists x)P(x,y)$, is also true.
Now, in the antecedent let $\alpha$ be such that $(\forall y)P(\alpha,y)$, and by applying syntactic rules for quantifiers and negation see that the consequent is equivalent to $(\exists y)(\forall x)\neg P(x,y)$. Next, in the consequent let $\beta$ be such that $(\forall x)\neg P(x,\beta)$, so in particular it follows that $\neg P(\alpha,\beta)$, and $\beta$ warrants the truth of $(\exists y)\neg P(\alpha,y)$, which is equivalent to $\neg (\forall y)P(\alpha,y)$, contradicting the earlier hypothesis $(\forall y)P(\alpha,y)$.