This is a free translation, paraphrase this article in 2006, which I came across about mailinator.com, authored by Paul Tyma. References Habré found, let him be translated!
So, you have a great idea.
Over the years, I started a small pile of projects and created several companies. For example Mailinator, Preemptive Solutions, Inc., and Classhat. In fact, I've done them much, much more, but most have heard about except that my grandmother. That's because it's mostly been a stupid idea, unsuccessful attempts to do something, or too ambitious undertaking, so I can bring them to release. Learn from my mistakes, I am now every new idea is checked against the certain set of rules to weed out the good ideas from the bad. Or those that seem to me those.
Do not forget, these rules for kanonіchnogo startup - when the employee is only you and your friend, and money - on beer and chips. If you have in store extra couple million lying around - you can do magic without these tips.
And here is the rule:
1) If there is no business model - it is a hobby, not a startup. I like to optimize the code. When Java first came, I wrote to her optimizer. I tried for months on this somehow make money, but it turned out people do not really steamed about the performance of their programs. And especially not willing to pay money for it. At least until a trained team of "salespeople" proves the importance and universal benefit from the acquisition of our product. In most cases, the ideas come from the environment in which I work, from the activities that I love. Only a few of them have a real business model where the consumer will bring you money. I have nothing against a hobby, but a business need to build on the fact that among other things generates income, not just the joy of hobbies you love.
2) The best ideas - those that allow your customers to earn. If you can say something like "With our product, you earn X% money" (where X - is greater than zero, even if not much) - a wonderful idea. Please note, not "save X% of money." Earn. The difference is greater than it seems. Save, of course, great, but compare:
"Give us $ 100 and you have a year will save 10 rubles.»
"Use our services and earn 5% more (and pay us a little bit of these 5%). Or do not use it and lose money. »
What would you have chosen?
3) It is best to work on business-to-business customer (you're here on the left). Yes, you will most likely be a firm that works with companies that provide the service your customers. If you are working directly with customers, get a load of a bunch of complexities. First, you will need to convince the client to pay for your services, and who likes to pay? Second, welcome to hell Support. You will always have a bunch of unhappy users, if you do not deal with support, and you can hardly - it is expensive and requires a bunch of human resources. If you have plans to do a paid service - do support system in the first place. Third, the more difficult to carry out many small transactions than a few large, you will need to develop the infrastructure. 1000 times more difficult to sell the service for $ 10 than 10 times the service for $ 1,000.
However, if you just provide services to other companies, it is also quite good. Large markets are difficult to control, and you can find their niche due to the rapid growth, or, for example, in the optimization of business processes serious companies.
4) If you choose to work directly with clients - pay attention to possible sources of income, not coming from the client. Make your service free of charge. You're thinking that I hit his head, but look here: a free service, so we can not guarantee a serious tech support or special stability, and it means - to reduce their costs. On the other hand free service paid more popular analog and has a chance to cause a viral effect. Now think about the profits from advertising and affiliate programs - voila, a free service that making a profit!
5) Sometimes remembers the old ideas. Why are all so fond of AJAX? Because with it you can do things that previously could not be done technically. Due to the advancement old ideas can sparkle with new colors.
When released Doom, has long existed 3D graphics. But the guys from ID the first to notice that the PC became fast enough to be able to make a shooter and play in real time. Technology stepped forward.
Every time something comes up or greatly cheaper - it is an occasion to reconsider their ideas. Traffic fell - there was video services. There were also cheaper smartphones - there was a bunch of attachments thereto. So always capture the moment.
6) Come up with something to earn income over and over again. Advertising on the site and all that - it's a simple way. But what if your product - application? Why did Microsoft change the details of the format with each release of Microsoft Word? Yes, the new chips, but in the end it's just a text, why not open older versions of documents made in the new? Just because Microsoft wanted. They want to be when you can not open another document, you went out and bought a new Word. They want to earn income over and over again.
From the outside it may look like a blatant greed. Have you heard about the "subscription to the software"? In fact, of course, "Software for which we will continually take money from you." If the products are not "wear out", no one would have bought new. However, if giving money to the user will see and feel the improvements, then everyone will be happy.
Still believe that your idea is good? Do not rush, there is the second part.
The second part of the tips can be found here (english).
Error please report to the PM, will correct as soon as I wake up. All've read here - thank you!