Sunday, 16 May 2010

minimal product

i agree with the lean startup gang of a minimal viable product. I think I understand it the same way as they explain it. The most important thing is the speed of learning.

have an idea? qualify it. Get feedback. Adjust. Build a prototype. Get feedback. Adjust. After this build a first version and do a 'soft launch' i.e. you actually have a working site... and you get real user feedback but you are not yet spending money on scaling up.

This makes sense to me. You've done everything possible to build the right thing but you don't truly know until its out there being used whether its correct or not yet. Then you adjust again and when you get that important survey feedback of 'I'd be disappointed without this product' and other solid reassurances of your products value, then you push hard with marketing and scaling up.

The alternative sounds like a nightmare in comparison. Have an idea... go into stealth mode for years while secretly building a monster and then when you are ready to reveal it to the world it turns out that no-one wants it. That's scary. So far we've spent 6 months on this project and it has gone from research, qualifying, specification to near completion. We can still be wrong... but it won't be as bad as having spent years on it!

I would like to see some better techniques for how you qualify something e.g. what really is a scientific method to qualifying an idea?

The first step can be just opinion. Ask lots of people in open discussion. When we started this project we tried brainstorming and it worked well. However, there is a technique to it. Get people together but don't tell them the solution and ask for an opinion on it. Give them the problem and ask them what their own solution would be. Run through multiple scenarios and only then introduce your technology in its simplest form and get opinions on how it solves that problem. Don't lead them to the answer which conveniently happens to be your technology.

At this point, I remember that the reason my co-founder and I are working on this project together is because he is the guy who will brashly go out and talk to pretty much everyone. Random people in the street, the guy on the table next to us, the girl at the bar, the guy on the train. He'll take over a dinner table to talk about the idea. That is why we work together and why I've also kept faith with him during the parts of the project where he could not contribute or was otherwise tied up.

In a normal startup, I think the developers rule. They will be doing the coding, they will produce the software and so what does someone from a Sales background do, during this time? Personally, I think there is always work to do in terms of marketing, which should be 50% or more of the project. Developers and Sales are opposite extremes of the personality scale in some ways however, if you can make it work then it will help you qualify your product much faster.

Its after this initial brainstorming that I'm wondering whether there is a more structured method of qualifying the idea. Maybe if you take a sample of people and almost all of them say 'That is brilliant' then you can do a calculation for the rest of the population. However, what if you work in silicon valley? Isn't your sample then flawed because even the guys working at starbucks will be saturated with dotcom news and ideas. So will that lead to your idea being slightly curved by the geek factor and then not hitting on what the mainstream want? It is well worth spending months qualifying something before any coding in my mind. Until you are sure.


Another step can be to do an interactive wireframe and by that I mean something that actually has some surface level functionality but takes a week to build at most. Wireframing tools need to up their game a bit more. They've got a lot better but they need an easy way of building in front end interaction.

Also, the wireframe gives you a first go at laying out the flow of the site. The user experience part is essential. I remember carefully thinking through parts of the user experience and changing the wireframes again and again and I still did not get it all right. Yet the wireframe I started with and what I ended up with were entirely different. Its shocking how much it changed the specification. I can't say it enough... use interactive wireframes to design the site.

So then you use those wireframes with users and see how they react. Layout the home page on the wireframe and see if they get what it is. Funnily enough, the wireframes really helped communicate among the team. I also realised that team members had a different perception on how this would work from my own ideas. Glad I did those damn wireframes, we could have been building a porn site by now otherwise.

Then what? I don't know. It would be nice if there was more material on how to keep the feedback/qualifying side of things running smoothly. I guess we keep talking to people. Everytime I head out with co-founder to talk to people, I learn something new. The other night I went to the toilet in this pub, came out and he's chatting to a Danish girl about the idea. Her experiences of making friends was different but it was definitely good feedback. She makes friends through other friends. Her interest was piqued when I explained how it could bring a group of Danish people in London together . Recently, having refined the pitch, people have been saying 'wow, that is amazing, when is it going to be finished?' Its when they ask that question that I stop feeling amazing.

Just-get-to-launch-please-


2 comments:

  1. here is a nice idea for startup research. Build persona profiles. I'm not 100% sure about this especially if you start to 'make up' persona's but otherwise I was thinking along these lines earlier in the project. Good idea just to try and imagine it as well but written persona's help
    http://www.webcredible.co.uk/user-friendly-resources/web-usability/personas.shtml

    ReplyDelete
  2. aha, another idea to add to 'how to qualify an idea early' would be to create the front page (just the design) and put it up on a site like: http://fivesecondtest.com/

    then you can get opinions on if they get the idea. In fact, don't even tell them which is your idea. Put it up and ask them which of the ideas was instantly compelling. something along these lines.

    ReplyDelete