This is a really great paragraph. I think the part that I get hung up on is right at the end - how does one actually go about finding an unfulfilled niche? It seems like they are kind of difficult to find almost by virtue of them being unfulfilled.
The niches I have found always come by accident: "hey, we notice you have this product that's close to what we need. Can you build us a version that does xxx" where xxx is some trivial change that results in a product I can sell to others in that niche. Or "we know this has to be possible and it shouldn't be difficult, but we don't have the skills do it. We'll buy xx units if you can build it for us."
The second-order problem is that often even if you know the need exists and you have a customer, without good knowledge of the domain, it can be hard to figure out how to reach other customers (they may not be online much) and make more sales.
I had a site that would consistently rank in the top 5-10 for google queries, yet I had fewer than 10 hits/day for those queries.
Another issue I've come across is that when you're approached by someone with a specific need, it can be hard to decide if you've found a new niche or just a custom design for one customer.
None of this is to suggest that you shouldn't try, but that it can be harder than it seems at first glance.
Still much much easier than finding a 1 in million unicorn idea and getting VC and getting users.
Judging from number of successful companies in both categories, it's several orders of magnitude easier (which is obvious).
What it does is add a common feature found in marketing automation/CRM systems to Gmail, a massively used platform.
It essentially unbundled a feature from a specialized product and made it available as on add-on on a prosumer product.
If you look at large platforms like Gmail, you can probably find isomorphic ideas.
After sorting through a bunch of junk about making an Excel wish list, here are some pages about suggested Excel improvements:
A previous company I worked at made YouTube videos for content marketing purposes. For some reason that no one could explain by the time I started there, the first couple videos in the account were random, basic Excel tutorials. Nothing fancy. Just things like using the Cell Formatting dropdown or doing a VLOOKUP.
Those videos were 5+ years old by the time I left. And in their worst week, a single one of those Excel videos got more views than the entire view count of every non-Excel video combined. It was crazy.
Edit: I should've read giarc's comment below! We're essentially saying the same thing.
I think that they are incredibly difficult to find and most people (me included) can't identify "problems" in their day to day that could be improved by something - in hindsight - relatively simple.
I'm constantly asking myself: where are the pain-points in my daily interactions with computers/software etc.
So far there's nothing and I know that can't be true.
Otherwise the best advice I'd give for finding problems is to start making a log of every time you get annoyed or frustrated at something. Your package didn't get delivered on time, you realize you have no milk when you're in the middle of cooking a meal, a driver on the highway is endangering people, your phone is too big for your pocket, radio stations all seem to play commercials at the exact same time, your home ran out of heating oil and no one will deliver until Monday, your hard drive failed and you realized your backup program wasn't including a new folder of vacation photos, even though you make an appointment at the barber/doctor every time you show up on time you show up you still wait 15-20 minutes, there's no Mexican restaurants that deliver to your house, you never know how your child is doing in school until they get their report card at the end of the semester.
There are dozens of tiny annoyances that we run into every day. Some have easy solutions, some less so, but they're there if you know what to look for.
Try to think of it in terms of outcomes.
What are you trying to do with computers/software, and how could the outcome be massively improved/democratized?
Can you do basic scripting? Can you do advanced scripting, with such revolutionary capabilities as calling out to external APIs for stuff like OCR?
Then rejoice! As you have the potential to make your very own billion dollar business! Just leverage those skills for process automation, market it as "software robots", leverage the existing perception of "robot replaces human" to make the recurring $10k/year cost for your automation script seem like a steal compared to the labor costs of a human, and through in some AI references for a bonus multiplier on your valuation potential.
Process improvement and automation has been a thing for ages. Marketing it as "software robots" is much newer. Turns out that particular phrase resonates really well with the target market.
I'm kind of talking out of my ass because I've never built a business but I've seen some terrible software that people are forced to use and been in many company-specific or industry-specific situations so I have buttloads of ideas. Most of my ideas come from hearing those pain-points or having my own pain points and searching for the right software and not being able to find it (or not finding what I need because I don't know the right way to look for it, which is another thing that should be a thing which is like, how do you figure out what kind of product you actually need?)
Start with things you already have some domain knowledge of (better even, try to find unrelated things you know about that overlap in interesting ways and where few people have domain knowledge of both) ;
find what problems/frustrations practitioners in that field have and do your best to solve one of them in a very immediate, low cost of adoption way;
fail fast .