NashCoding Yet Another Artificial Intelligence Blog

10Apr/110

Building a Startup – Part 1 – Introduction

I've joined with David Fogel and Yanon Volcani to build EffectCheck. This post is the first in a series detailing how we're building this startup from the ground up.

What is EffectCheck?

EffectCheck is an algorithm that draws on decades of clinical psychology research to identify the subliminal emotional impact of your word choice.

Lots of words have subconscious effects on us. For instance, if you're walking down the street and you see a sign with the word "ELDERLY" written on it, studies1 show people actually start walking slower without realizing it. Alternatively, consider responding to someone telling a great joke. You can say "That joke is hilarious" or "That joke kills me" -- both mean the same thing, the context is the same, but subconsciously the word "kills" actually raises your stress levels. EffectCheck measures these subliminal effects and helps people discover how their writing likely makes readers feel independent of context.

Who could use it?

Anyone who wants to understand the emotions their readers are going to feel. Some examples:

  • Advertisers may want potential customers to feel more confident so the customers buy more merchandise.
  • Defense attorneys writing their closing arguments may want the jury to feel compassionate towards the defendant.
  • Politicians giving a speech may want to charge the audience by raising anxiety and hostility.

There's clearly a wide range of use cases and EffectCheck can be customized to specific domains given enough training samples.

What will be covered in this series?

I'm new to web development, so I'm hoping this will serve as a semi-roadmap for people looking to build their own startup. I'm specifically focusing on people who have cool algorithms or applications that they've written and want to create a startup around them. Some of the topics I'm planning on covering:

  1. Why I chose .NET
  2. Running a 99Designs logo contest
  3. Building a developer API platform
  4. A/B testing, funnel analysis and metrics

There will be a lot more to come also, but that's all I've dealt with so far. With the exception of this intro post and the "Why I chose .NET" post, I'll try to keep everything technical or instructive; I don't want this blog to become about my random musings and opinions. I also will not be covering any of the implementation details behind the EffectCheck algorithm (trade secrets and NDAs). Everything else is fair game, so if you're looking to build a startup, I hope you find this series helpful!

Footnotes
  1. For a better discussion of this experiment, see Stumbling on Happiness. It's a great book in general, highly recommended.
Comments (0) Trackbacks (0)

No comments yet.


Leave a comment

 

Trackbacks are disabled.