Ask HN: How to do market research for product?

really good info

As you already have a niche in mind (restaurants that are already on delivery apps), you can make a list and try to approach them.

How I'd do that (consider it a brainstorm):

- I'd build a simple prototype

- If I have friends in the niche, I'd approach them and ask for feedback

- I'll actually ask them who is managing their menus and marketing. It makes sense to try finding agencies (these could be graphic or print agencies) that print the menus. They can get you in touch with restaurant owners and maybe they know something about keeping menus up to date.

- I'd write down what their struggles are. For example - expensive to update high-quality printed menus, challenge to maintain onsite digital menu, challenge to match current availability with online platforms, etc.

- See if any of the issues they're referring to, does actually relate to the hypothesis you've built. If you're not happy with the result, iterate.

- Play around with use cases - list a couple of use cases (Client orders online, client cancels online, client orders on the phone, order is paid but no availability, out of stock, etc). Observe the business processes surrounding these use cases in the restaurants and see how your solution (or hypothesis) might fit there.

- If you don't feel threatened by the delivery apps, try approaching them. They are usually pretty well-connected with the restaurants and know a lot of their pain points.

- If you can afford it, spend some time in real observations (I'd do that by sitting around a friend who has a restaurant and accepts orders).

If you have a prototype that you believe can be validated, try it with a restaurant. If you're unsure how to approach a restaurant owner, manager or staff, just go there and order some food. Then present yourself and express curiosity. My experience is that if they are not busy and if you don't ask for money, they could be talkative. The more you order, the more talkative they get.

Be very careful what you promise for free to the restaurants. They might ask you to build them a website or help with Windows update.

Good luck! Hope you got some insights.

I feel like a weird industry has popped up that makes you feel like you can validate everything in advance. I think this comes from Product Managers in large, risk-averse companies that excessively try to minimize mistakes.

However, if you want to start a startup, you aren't optimizing for mistake minimization! You are figuring out if anyone cares as quickly as possible. You will only get true validation of this by seeing if people use your thing, and then iterating from there.

My advice would be - you have to _both_ talk to people and _at the same time_ create whatever your MVP is to launch as fast as possible. It might be a spreadsheet. It might be you doing stuff manually for them before building the software. Get something into their hands as quickly as possible.

There is only one reliable way to do market research, and that is to sell the product. Find a restaurant with an outdated website. Go into the restaurant and take a paper menu and create a website with the updated menu, make the website responsive, if you can't upload the information to Yelp, screenshot the Yelp page and replace the prices and photos with Microsoft Paint.

Walk into the restaurant and sell it to the owner. If he buys it, congratulations! Now you have to go build it.

This is a really useful post thank you