There are many programming languages, and very few of them enjoy true popularity. The others are stuck in niches. Sometimes a languages suddenly gets some attention among software developers because some popular application is written in it, or because it supports a programming paradigm which became popular and after some time it returns into near-to-obscurity. The fact that Ada is not popular is unfortunate but there are many, many languages sharing this fate, and it does not mean that something went terribly wrong.
Surely there are some factors working against Ada: To many its wordy, Pascal-like syntax just doesn’t feel cool. Many people are OK with writing an application that just sort-of works and fixing the bugs (or rewriting the whole crap) later. Ada appeals to the opposite mindset. That its implementations lack automatic memory management means it will probably not be able to compete in areas dominated by languages like Java and C#.
That said, I think there are areas where Ada can show its strengths. That it is possible to prove the correctness of a program instead of writing tests for a lot of cases, hoping that you covered all important cases is something I would call a cool feature. But it is hard to do in practice, so formal verification may be restricted to areas where correctness is extremely important, like when human lives depend on it.
I think there are two things Ada programmers can do: Writing Ada versions of popular applications or libraries to demonstrate that it’s no less useful than other languages, or trying to write some ‘killer application’. The first is arguably easier to do. Choosing between the first and the second is up to each individual programmer.