how to validate an idea

Showing 1-3 of 3 messages
how to validate an idea karan 1/25/12 8:27 AM
Hello group

i am very new to lean startup group

I have learnt a lot so far and am very thankful to all members

I am running a software dev company and have a lot going on with services , products , support

I have 2 new ideas which - I feel are better than most of my ideas so far

Since we have very little free time now - how can we possibly validate our ideas ??

one more problem is that we have not worked on any b2c idea so far and both these ideas are in the b2c area

I shall look forward to your inputs .

Thanks and best regards

Karan Ahuja

Re: how to validate an idea tRavIs 1/25/12 7:25 PM
Karan,

The best way to validate an idea is through customer interviews.
List the top three problems you are trying to solve for each idea.
Then test them with your market by asking if the problems resonate and
how they solve them today.
Get your interviewees to rank the severity of your assumptions and
possibly provide issues they are struggling with you have not
considered.
The goal is understanding the customer pain level and existing
alternatives.
Is a potential solution something they must have, would be nice to
have or simply do not need?
If it is something they must have and are actively seeking a solution,
you have validated the problem.

tRavIs McCutcheon
Re: how to validate an idea Michael Norton 1/28/12 5:44 AM
Customer Interviews are a good means of validating a problem, but not as effective for identifying a solution or validating a new and unique idea. Most cannot answer questions about interactions or usability of a system we've never previously experienced, nor can most imagine a truly new or creative solution to a familiar problem. Few if any mobile users in the 1990s knew they wanted a flat, glass-front phone with icons and a full keyboard. Had you asked, customers would have told you their problems were the quality of the existing hardware. Many a company spent a good deal of money creating and validating better versions of the same hardware. The companies that made a real impact spent money creating entirely new phone designs.

To validate a product idea, you need to observe people interacting with the product. But you don't necessarily need to build the concept in order to achieve this. Start with a key fundamental aspect of your solution and figure out a way to verify that people see benefit or would engage. Let's say, for example, that I believe people would pre-order their food while waiting to be seated at my very busy restaurant and I want to build a mobile app and kiosks to enable this behavior. I might start with a simple sign in the lobby that says "Want your food faster? Contact the hostess to place your order now." If this doesn't get any activity, then I may try another version of the sign. If that doesn't get any activity, I need not build anything more.

Once I've verified people will pre-order, then I might create a simple web form that prompts for name and order (free-form text), place that in the lobby and put some menus next to it to see if folks would actually use a kiosk system.

Continue to make small adjustments and see if these have positive or negative impact on customer interaction. Count the number of times people return to the system, not just the number of people who use it once. Focus on those who return and work to improve their experience. The others will follow as the system improves.

- Doc