I have a bit of perspective on this since I make decisions to hire people to do this kind of work, so I'll share how I think about this.
In general, being an OR professional typically requires a PhD, ideally in something industrially relevant. This is by no means always the case, but the field is not yet at a level of user-friendliness that allows people who don't deeply understand the fundamentals to these jobs well, unless the job is just very easy anyway. This has nothing to do with how bright someone is. Optimisation professionals simply need to know too much before they can be useful, and accumulating that kind of experience takes time.
Our world is a big place and there are very diverse opportunities. Mine is merely one opinion, but I would personally not consider an entry-level professional without a PhD for an optimisation position, simply because training them to do the things I'm not smart enough to do myself would take years. For perspective, we spend 6-9 months training people with a PhD at my company (although admittedly we do high-tech stuff).
Don't get me wrong, it's not that a PhD gets people automatically hired, but it will get them an intro phone call if their field of study is relevant to what we need at our company at the time.
Note that all this applies to young professionals. When it comes to someone with work experience in the field I don't care what their academic credentials are, as long as they can do the job.
The more subtle factor here is actually something common throughout hiring decision-making in technology: "why are we hiring someone?". There are exactly three reasons:
- We need more manpower on this.
- We have no idea how to do this so we need someone who does.
- We know what we're doing but we lack specialist know-how in this specific thing.
With all this in mind, your first decision should be what aspect of the work appeals the most to you. Do you enjoy solver technology, metaheuristics, or modelling? If you pick one of those, which sub-topic within them do you find most interesting?
The second train of thought is practicality. Among the things that you do find interesting, which of them are used the most in practice, or will be in 5 years' time?
Your third decision is what kind of professional you want to be. Do you want to be a generalist and work at a small consultancy, or do you want to be the absolute best at one thing and get hired to do the really high-tech stuff? The former ensures there will always be some job you can take, while the latter means that maybe 100 people might ever be interested in hiring you and would pay big money for your skills, but it might be years before you find such a position.
Once you decide what's an acceptable trade-off between these factors, you have your goal. Once you do, figuring out how to go about it in practice is the simpler part. Look for a PhD under a supervisor who knows what they're doing, apply for relevant jobs/internships to get hands-on experience, or start coding/contributing to an open-source optimisation codebase.