How to work on a new Product idea?

Question

Hi -

I was thinking about a software product. There are products that do similar tasks but as far as my market research goes, there are no specific products that this one replaces. So I guess I have a good case. How ever, I wanted to know the answers from some one who successfully converted a product idea into a product (and made money of course)

1. How do I document the idea that gets the interest of a potential customer? Should it be technical or should it be based on a use case?

2. Do we always need a prototype?

3. Should I make the product and then pitch for customers?

4. Well the problem is I have the idea but just like any other product, It has to come with 1. Licensing model 2. Security system 3. Documentation. I being the only person right now, implementing all these would increase my time to market. Does any one have hints on how to tackle this problem?

Any hints and pointers are appreciated.

SK

Answer 1

Perhaps you should first do some research with your potential customers and see if they would be interested in said future product.

After getting favorable feedback I would focus more on the product itself and when you think that it is definitely something that will come to bear, THEN start thinking about your licensing, security system, etc.

Answer 2

• Market Research
  • Competitors
  • User Base
  • Window of oppertunity
  • Product type
  • Groundbreaking
  • Some existing products
  • Advanced solutions exist
• Technology
  • Current State
  • Trends
  • Failures
• Proof-of-concept
  • Design
  • Time-to-market
  • Do you have tools you (will) need?
• Cost
  • Profitability
  • How long can you go w/o hitting break-even?
• Why will the idea never ever succeed? (Play devil's advocate.)

These are listed in no particular order of importance. A surprisingly large number of questions can be answered if you know your way around the Internet.

Answer 3

Whenever you find a hole in a market, the question is: Is there a hole because you've found an opportunity, or is there a hole because no one wants it, or because the hole is too small to be interesting.

Usually it's the former if you're a one-man shop. Because even a tiny niche is enough to drive a one-man shop.

First thing is to gather 6 people that will use the software when it's done. Probably you'll end up giving it to them for free for all their hard work in debugging it, but you need some people to say they'll buy it.

Further tips:

1. Make sure you can give it away to anyone, for free. If you can't give it away, you certainly cannot charge for it.
2. It's OK if your first idea fails. That's normal; keep going.
3. Be careful with advice (like this!). Evaluate all the advice you get and make your own choices. Don't be afraid to change your mind in light of new evidence.
4. This is a great time to start a new company; use these tips.
5. Here are five ways to listen to customers and hear what they really want. Biggest rule: Listen to their root pain and complaints, not their specific features.
6. Don't just go with the consensus -- usually the consensus is bad when it comes to creative work.
7. Use a free system like Uservoice to track features and bugs. Do this even in your planning phase. Don't miss a chance to collect data!

Answer 4

SK,
I think there are several things you probably need to do.

• First, you need to build a prototype. I don't think you should worry about the business details until you have a product to sell.
• Second, you need to do some reading on the business of software. There are number of good books on the subject, and there are plenty of resources on-line that can give you some idea of how to start your own business.
• Third, you need to network. You need to do this for 2 reasons. One is to meet potential customers. The other is to meet potential partners (either on the coding side or on the business/vc side). It's extremely difficult to start a business alone (it's not "Steve Jobs" or "Bill Gates", it's "Steve Jobs AND Steve Wozniak" or "Bill Gates AND Paul Allen") and having a partner will prevent you from having to do EVERYTHING.

But really, it boils down to, build the product first, before you go trying to sell it to people. Would you give money to someone just for an idea, or would you rather see something that works on some level and demonstrates that you know what you're doing?

Answer 5

You will inevitably need a prototype of some kind. You are presuming that your market research is thorough and that can never be tested enough, so keep up that investigation. Next step is to branch into two efforts

1. design enough of the user experience that you can present it to some real users and get some feedback. Make that feedback include what they would pay for it, what is missing and what else they think it is like - that will round out your market research
2. start thinking about your architecture and a development plan. If your idea has legs and there's a market you'll need to have that up your sleeve so when you start to do some real work you have an idea of where to start. Also if you decide that it is compelling enough to go and get someone else to back you, then you'll need to provide an estimate of how much effort is involved. The sooner you have a feel for that the better.

Your prototype can be in PowerPoint, you don't have to have a lot of code in order to get early feedback. The most common mistake people make is presuming that there is a market and they have crackled the problem. It's almost never true and you won't know until you start talking to potential customers. Don't worry about the technical end of licensing yet, you can deal with that later. In any case you may decide that you want a different delivery or pricing model than the one you are setting out with, so the later you leave the licensing the better - or at least be prepared to be flexible about it.

Good luck!

Answer 6

1. It's very unlikely a customer will be excited about some white paper or document regarding your product. Implementing an idea is hard, as you point out in your post. Lots of people fail at this step. I think you just want to identify the problems that your application is trying to solve, and communicate that to the user. Generally, videos, screen shots, screen casts are the best way to reach the most number of people. Documentation generally isn't catchy enough for most people to notice.

2. I think so. You need to actually create a pretty functional version of your product. The best place to start is to list what problem(s) your product wants to solve. Then build the minimum amount of software you need to solve those problems. The end result has to be enough to be compelling for the user to use right away.

3. Yes. In my experience building a product first is pretty much the only way. It is really hard for a garage venture to get people excited about something without actually having them see the product in use. It can be done if you are well known in the industry, and have a great track record. But, this is pretty rare.

4. The main thing is to have a product that your potential customers can use right away. That's how you build creditability, and word out mouth following, which if you are just a one man operation is your main marketing too. You can worry about the other stuff later.

To list some of the other stuff you mention.

Documentation: This is not very necessary at this stage. I haven't encountered too many customers for whom this is a major stumbling block to buying your product.

Security System: It depends what you mean here. If you are worried about people stealing your software. I wouldn't worry too much about it in the beginning. The revenue you are likely to lose because of this is not going to be huge in the beginning. If you are talking about user security, then that is important, especially if you are building some banking or commercial software.

Licensing model: This is also not too important to get right in the beginning, because most likely the model you start out with will have to be tweaked over time. The reality is that there is no way to really gauge what your demand curve will look like. You have to experiment. Pick any model that sounds reasonable, and then adjust it over time.