Ask HN: How do you come up with ideas for a website/product?

Websites which irritate you: clue to UX design &quot;do it better&quot;<p>Websites which are incomplete: market demand proven, &quot;do it better&quot;<p>Websites off to one side: find your niche<p>Do one thing, but do it well.<p>Do one thing, but be comprehensive with pointers onward.<p>Be responsive to feedback.<p>Provide a service to people in your niche: finding websites with bad copy? write a simple tool to fix simple copy mistakes and give it away on github but sell the service of doing &quot;moar&quot; but applied to your niche.<p>Go to conferences, and talk to people in your niche and pick up on <i>gee I wish that x was done better</i> signals

Often my ideas have something to do with an inefficiency I&#x27;ve seen, or a nagging problem that doesn&#x27;t seem to get the attention that it deserves.<p>Watch lightning in slow motion sometime. Many times before I come up with an idea, my mind is racing and seems to send out branching tendrils through resources, problems, techniques, news, people, previous ideas, etc. Like step-leaders in lightning, many alternate paths and branches are evaluated and activated. As this network expands, a solid connection forms between a problem and a solution, and BANG! An idea forms. Sometimes with many important details baked in.<p>I focus in on this active path, and debate the merits and ramifications and potential pitfalls. Sometimes this idea is valuable and new. Many of my ideas have nothing that would connect them to anything I&#x27;ve worked on before.<p>Often these Ideas will get lost if I do not write them down, so I always endeavor to do that. Reading through these old ideas is a good way to re-activate these pathways of invention again and generate new ideas.

It&#x27;s worth noting that if you look at the big tech companies (MSFT, Google, Apple, Facebook, et al) none of them started with a particularly original idea. They just executed far better than the incumbents.

I wrote about this a few years ago and you might find it useful [0].<p>It is really hard to come up with good ideas.<p>0. <a href="https:&#x2F;&#x2F;www.tillett.info&#x2F;2015&#x2F;08&#x2F;30&#x2F;ideas-are-not-cheap&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.tillett.info&#x2F;2015&#x2F;08&#x2F;30&#x2F;ideas-are-not-cheap&#x2F;</a>

Along the lines of build something you need, think about your life and the way in which it could be better. Do some research on your ideas, if a product exists then you’ve found a way to improve your life. If it doesn’t exist, you have an idea. Either way, you win.

Another way of looking at it is to focus on the process and see the idea as starting point.<p>Don&#x27;t be discouraged if there&#x27;s an existing solution as you are likely to end up in a very different place from where you started.

Ideas don&#x27;t matter much. As you&#x27;ve noted, most have been made many times. Do something you&#x27;d enjoy to work on and do it differently. Don&#x27;t use the same tools for example.

<i>Many people have said: build something you need. It seems like all my ideas have already been done.</i><p>For me at least this has been accurate (building something I needed). Tesults, <a href="https:&#x2F;&#x2F;www.tesults.com" rel="nofollow">https:&#x2F;&#x2F;www.tesults.com</a>, is an automated test results data storage&#x2F;reporting web application. As a developer I had to implement a custom solution to do this for three different teams across two companies (employers) over a three year period. After the third time I noticed the pattern. I also noticed how we used it everyday, throughout the day in our dev team so it seemed like a fairly good problem to tackle in a general way, not least of all I thought that whatever was created could be used at my next job.<p>When you say your ideas have already been done are you sure they have been done exactly in the way you&#x27;re thinking of, or done as well&#x2F;polished as you would like? With the problem Tesults solves there are open-source and paid software alternatives out there that provides the building blocks to do something similar. Alternatively you could build something yourself like I previously did. Some people would definitely choose one of those two options. However at my job none of these options were great due to missing features we needed and maintenance concerns. As an employee this is was not the main part of my job so it was quite annoying not having something available to just handle this, I didn&#x27;t want to manage the web app or database and my employer didn&#x27;t want me to either really. So yes others did have the same general idea but they had not executed as well as I would have liked. It wasn&#x27;t a slam dunk though, after the initial release it took over a year to get the first couple of paying customers (had several dozens of free ones) but that was because earlier iterations lacked polish and were missing key features (such as logs, screen capture, file uploads). Also I didn&#x27;t have someone helping with service or marketing before.<p>I really think execution matters above all else and continuing to iterate is important. Tesults needs continuous work, it&#x27;s updated every day, usually multiple times and there are years of work in the pipeline to make sure it remains the best way to report results data. If it wasn&#x27;t for my seeing the problem in the wild as an employee I would never have done it, the idea would never have occurred to me sitting at home, I needed to have worked in the domain and I needed to be a domain expert. I think if you are a younger person you can perhaps think of cool stuff for you and your friends (b2c) but I&#x27;m an older person now with thoughts dominated by business issues and around being more productive. That and diapers (not my own).