I'm told better to build something at least a few people love, rather than something many people simply like, as it's easier to scale a loved product to the masses than it is to get even a few people to adopt one that they merely like.
My concern regarding this approach is that I worry my product may not have even the potential to become mainstream (i.e. ability grow out of it's user base of early adopters to the early majority).
What are some good ways to gauge whether a product that's loved (albeit by few) has the potential to grow outside of it's initial (somewhat niche) market?
The easiest answer to this is something a mentor told me a few years ago when I was working on a niche app....she said "do you honestly feel like you could be passionate about this for the next 10 years"? That may not sound like it's relevant but it is...honestly that's the only thing that matters when starting a compay (besides the obvious of is this valuable)...That's the first thing to figure out. Then when I answered and said maybe, she said "do you think that this is the biggest thing you could be doing with your time? Why stop at helping a few thousand people, why not pour 10 years of your life into something that can help millions".....so reframe your question, is my product only helping a few people? Is there something else I could be doing that will help more people? If the answers are obvious then you know what you need to do...back to the drawing board.
Niche is a good thing! It'll help you stand out. I would do a lot of cold emailing and targeting to that specific niche to dominate the market.
You can use something like http://www.growthok.com to help you find targeted contact information. Hope that helps!
In my experience every product or feature finds traction with 1%-5% of users - indefinite retention, express strong satisfaction in surveys etc. For paid products the number is usually much lower. The questions you must answer are:
1. What's special about these people? Don't assume they are "techies" or "early adopters" - these answers are too generic and often wrong. You need to reach out to your best users and interview them. If you do enough of these a profile will start emerging - single parents living in suburbs, <50 person companies in the transportation sector ...
2. Why do they love it? What job is it doing for them?
Once you have the answers to both you'd be in a good place to answer the question you're asking - is this a seed of a much larger market that's will grow significantly over the next few years.
The better question is: do you want to go mainstream?
If your startup is satisfying a niche, then your strategy might be to jump to other niches. If you're solving the needs of grocery shops with a small ERP, then you might jump to bodegas, to bakeries, to small coffee shops.
Going "mainstream" is totally different - it's the final step, after you've conquered all the niches that you can think of, where the product is "generic" enough to spread to dozens or hundreds of tangential markets. Most niche startups never even make this transition.
The real question, therefore, is not if you should go mainstream. It's: is your niche market (and the subsequent niches you'll conquer) big enough and will it grow over time? it's that simple. There's no worry about building a product only for grocery shops if they pay enough and there is enough of them. If you're not making money off of them as the market isn't big enough, trying to generalize features and "phase-shift" to the mainstream will not only not work but make you lose focus and turn your product to crap.
You can make all kinds of market estimations, but at the end of the day, has the niche market (or markets) paid you enough to survive, and, even if not, do you truly feel that you can capture more value from them (by adding new features and cool stuff) in the near future? If yes, viable. If not, run away.
Take "mainstream" out of your dictionary and mind and focus on niche market size and growth.
You can always scale up because you can always find some very related industry that can use your products/services with a minimal change.
For example, if you make specialized software for orthodontists and have captured the market, but have happy customers. Then, you should be able to add other types of dentists, maybe by creating a slightly different product that allows for their expanded types of services.