Hacker News new | past | comments | ask | show | jobs | submit login

That was a really great article. Here's a few notes from my experience with FormAPI.io:

> One of the best ways to do it is to watch people interacting with your prototypes without giving them any hints.

This is so true. My product has a lot of very complex features, but after watching some early users, I realized that people were struggling to figure out some of the most basic things - even just adding and removing fields. I had made the 'delete' icon only show when you hovered over the field in the sidebar, but I made it more obvious by always showing it when the field was selected. I also added another 'Delete' link to the field option in the right sidebar. And then I also added a welcome modal that explain how to add and delete fields with animated GIFs. (Including the "delete" and "backspace" keyboard shortcuts.) Now that there's 3 separate ways to delete fields, people figure it out almost immediately.

Even adding fields was hard to get right. My initial version just had click-and-drag, so you had to first click, then drag to adjust the field width. I saw that new customers were just clicking on the page, and nothing would happen. So I added support for single clicks as well, and now it adds a field at that position with a default width.

It's scary that I might not have realized these UX problems if I wasn't paying attention.

> Nowadays it’s easy to get amazing stock templates for less than $50 instead of spending thousands on a custom design.

I strongly recommend https://pixelarity.com. I've used 3 of their templates now, and I think they're very good. Some of them even have jekyll versions that you can download. http://unsplash.com is incredible for stock photos, and I have a subscription to https://thenounproject.com for vector icons. There are a lot of free icons out there, but the Noun Project has a ton of variety, and lots of things you can't find anywhere else.

> When doing the development yourself, it is easy to slip into a false productivity when you increase the amount of code without achieving your business goals. The most infamous examples are building tools instead of a product or overthinking a complicated architecture "for the future."

This is really tough, but I'm actively trying to avoid this. Not just the fun tools and side-projects you want to build, but even some feature requests from customers. Certain might end up taking weeks or months, and they'll over-complicate your product, or they just won't be a good investment of your time.

> New and trendy technology usually means bugs, breaking changes, immature tooling and lack of documentation. "Boring" mature tech will allow you to achieve the same goals much faster and make it easy to find developers for hire.

I used Rails and React, since I'm very familiar and productive with those. I was tempted to try Elixir, but I think it would have taken me so much longer to build an MVP. I have a friend who started building their startup with Elixir, but switched back to Rails for the better productivity.

> Depending on the country, your experience with these business aspects can range from mildly unpleasant to the absolute worst.

I've almost finished setting up my company with Stripe Atlas, and I can say that it has been the absolute best experience. I can't recommend it strongly enough. I'm going through the post-incorporation stuff now with UpCounsel, and everything is just so easy.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: