I'd say that "successful" is fairly personal for most projects, as your expectations, past, needs, and wants shape whether you would consider a project "successful" or not.
For example, the game "Among us" is an interesting case. The game was released two years before it really hit off like it has recently. One could argue that "Among us" is now very successful, but had the developers not kept the project maintained for the two years prior to its current popularity, it may have never been the hit that it is today. Whether one considers "Among us" successful during those first two years really depends on what factors they are looking at.
Monetarily, the game is certainly successful now, but in the two years prior, it likely didn't cover much more than development costs (if even that). However, I have no doubt the developers learned quite a bit from the experience, so from that perspective, it may have been a great success.
Ultimately, they continued to maintain the game for two years, so they obviously saw it as a worthwhile project during this time due to some factor.
All of that above is to say, a project can be successful or not based on what you are looking at. Personally, I view success as players enjoying the game, so if people seem to be enjoying it, then it is a success. I also think any project that pushes my skills forward is a success, even if I do not see how it may apply in the current moment. I have dozens of "failed" projects that I have later realized have been important stepping stones. If you are pursuing development as a career, then making enough money to sustain the career (and any expenses) is obviously an important factor that should be considered.
I cannot really help too much on the marketing stuff. I only have limited personal experience based on my own observations when it comes to marketing. Professionals who work in marketing or developers who have worked with said professionals probably will have a better idea.
As I mentioned before, I just post on my Twitter showing interesting aspects of what I am working on, and hope that it generates interest. I know that consistency is important based on my own observations, so if you are looking to market, that is something I would keep in mind. Additionally, I know that reaching a large audience is important for sales, as only a certain percent of whomever views your project will actually be interested enough to buy/download it. Knowing the audience you are targeting with your project so you can tailor your marketing towards the groups most likely to be interested is probably a good idea, but I have no idea where to find that sort of data, outside of just observing similar products and seeing how they do it.
