December 16, 2009 — If you combine Paul Graham's "make something people want" advice with Sean Ellis' product-market fit advice (you have product-market fit when you survey your users and at least 40% of them would be disappointed if your product disappeared tomorrow), you end up with a possibly even simpler, more specific piece of advice:

Make something 40% of your users must have

Your steps are then:

• 1. Make something people want.
• 2. Put it out there.
• 3. Survey your users. If less than 40% would be disappointed if your product disappeared, go back to step 1.

Only when you hit that 40% number(or something in that range) should you be comfortable that you've really made something people want.

Does this advice work? I think it would for 3 reasons.

PG and Sean Ellis know what they're talking about.

I made a list of my "must have" products and they are all largely successful. I suggest you try this too. It's a good exercise.

My List of Must Haves:

• Google Search
• Facebook
• Gmail
• Dropbox
• craigslist
• Windows
• Excel
• Twitter Search
• Firefox
• Chrome
• Wikipedia
• Amazon

• (More Technical Products)
• Git + Github
• LAMP Stack
• Ruby
• Notepad++
• Vim
• jQuery
• Firebug
• Web Developers Extension
• StackOverflow
• TechCrunch
• HackerNews
• Navicat

I've worked on a number of products over the past 3 years.

One of them I can tell you had a "I'd be disappointed if this disappeared" rate of over 40%. We sold that site.

All the others did not have that same "must-have" rate. We launched Jobpic this summer at Demo Day. People definitely wanted it. But we didn't get good product/market fit. If we had surveyed our users, I bet less than 10% of them would report being disappointed if Jobpic disappeared. Our options are to change the product to achieve better product/market fit, or go forward with an entirely new product that will be a must have.

I don't know if this advice will work. But I'm going to try it.

Startup advice can be both exhilarating and demoralizing.

On the plus side, good advice can drastically help you. At the same time, if it's really good advice that means two things:

• 1. This is how you should be doing things.
• 2. You were not doing things this way.

That can frustrating. I've spent a few years now in the space and to realize you've been doing certain things wrong for a few years is...well...painful.

But you laugh it off and keep chugging along.

• Thanks Nivi for the great Venture Hacks interview!
• Users/Customers refer to people who use your site regularly or buy from you. This is not "visitors". Generally a much lower percentage than 40% of visitors become users or customers. The 40% refers to the people who have made it through your funnel and have become users or customers.
• I used customers and users interchangeably. For non-tech businesses, you can just use "customer" each time.
• Thanks to Ben, Alex Andon, and Andrew Kitchell for feedback.
• Another piece of startup advice that didn't "click" until recently: Roelof Botha's 7 deadly sins advice.