Ask HN: How to identify niche fields that would flourish with the aid of tech?

[arcontethegreat]

The proliferation of data and computational methods have proved to be beneficial to a handful of fields (Computational "insert physical/social science here"). Rather than assuming that marrying tech with existing practices would solve everything, are there more optimal methodologies in identifying fields that would see an exponential boost in utility?

[tidenly]

One common issue I see with programmers is that they often think a knowledge of technology alone is enough to identify and fix problems in a field.

You actually need a domain expert, or to become one yourself, within some field to even begin to know what the current issues faced are, what technologies or products are already in use, and what kind of solution would be acceptable in the market. The balance then of finding the solution is actually much more 50/50 than a lot of tech people want to believe - its not just them sailing down on a cloud and delivering some perfect pre-made solution.

Usually I find these people through my friends or family. You basically have to sit down and interview them, and have a back and forth to find what kind of things would or wouldn't work in their industries. I don't think theres some magic bullet to finding all untapped markets.

[EdwardDiego]

All of those niche fields already have tech, often an incumbent company that started with MS Access and is struggling to scale.

Two niches I encountered, independent music schools, garden centres/nurseries.

So you're not going to bring a flourish just by bringing tech, they already have it. But can it be done better? Hell yes.

Can you make a business model out of doing it better?

...I don't know. In my country, that's a lot of start-ups that target agriculture, and deliver some good results, but then find out that farmers are risk adverse when it comes to new spending.

An old colleague developed fantastic software for managing hives, and then found that the beta testing apiarists who loved his software were even more risk adverse than farmers.

But yeah, if you want to target a niche, you need a partner who not only knows the problem domain, but the market/social network, and has good relationships, as sales will initially be word of mouth.

[kungito]

My experience is that you have to know someone in any other field and they will know of things that need to be automated

[mywittyname]

You also need to find a person in an appropriate field. Not every field is easy for startups to break into.

Example, my partner is a nurse, but works in an office capacity. We've co-developed a suite of tools that make her job so much easier/faster. These tools could probably justify a $100k per annual seat and a small client might buy 20 seats. Yet, we can't really productionize it and sell it because healthcare is basically impenetrable.

So if your goal is to sell a product, think through sales and marketing before diving into the code.

[pgwhalen]

100k per seat annually? Just to be clear, you’re saying that said software would cost roughly $100k/year/user? Is there any pure software product in the world priced that aggressively?

[matyasrichter]

This is very true. The key is to avoid the classic "introduce a problem you didn't have so that you could solve it with a product".

[fxtentacle]

Everything where there's a lot of people working on it. Tech tends to replace employees and the more get fired, the bigger the savings. We then call that leverage.

- customer support

- health/beauty/nutrition advice

- "content creation" ie blog post SEO spam

- tax/legal

- digitizing, managing, filing out paperwork

- supermarket cashiers

- government bureocracy of any kind

[exodust]

> supermarket cashiers

I don't use the self-checkouts anymore since they installed cameras on the screens at some of the supermarkets in my area. I don't care about normal CCTV in ceiling. But when the camera is 30cm from my face as I pack bags, it's literally a mechanical eye replacing the human eye of the cashier.

[pydry]

Find a cofounder who has been making customers happy in their field with crappy jury rigged automation (e.g. spreadsheets).

[Simon_O_Rourke]

The legal field, at least in Europe and England, is almost completely paper based. Like, for everything, from court dockets to lawyer's notes. However, it's so ingrained that I doubt it's open to change.