Ideas

Smart !

Super hot, sounds like an exposed version of Google search indexing and how digg does content associations statistics ( social clumping around topics ). I wonder if someone out there is already doing this, it seems like it would be.

Google has a lot of keyword analysis tools in the AdSense toolkits. They let you look at current searches for various words, etc. I'm not sure how much they do to correlate multiple words together, but you'd need to really differentiate yourself from that existing offering.

Google only makes a very narrow range of information available on keyword tracking, because a lot of it uses spidering technology that they utilize. Contextual spidering is a lot of what is at work with PageRank, but this would be making use of it in a much more precise way. Rather than just looking at what keywords appear together over a number of sites, we're also looking at their frequency and how they dilute as you reach the "boundaries" of your niche and hit another one. This is supremely awesome because it allows you to track trends fractally. ;)

Yep I like it. This is very solid.

This is really interesting. There are a lot of 'walled gardens' on the web, and I think it would be really useful to be able to identify them.

there are a couple of similar things out there, I've been poking about. most are overlays on a site for admin. check out Click Density, they are one of the better ones:

http://www.clickdens...Da75ICFQTWsgodSjX65A

This idea you are talking about is a bit different, but in a similar realm.

I like clickdensity's approach, and you're right, it is a similar realm. Site overlays are not something that should need to be utilized, though that may become different as this idea evolves. It's a good preview into how we would market this type of service, though. Thanks man.

Chalibaeus: check out http://labs.digg.com/ as well, the boys over at Digg Labs do some funky fun stuff, all ways of visualizing statistical data.

Do you have any econometric background? Would you use some form of commercially available statistical package and programming combination (SAS, SPSS, etc?), or do you have some sort of proprietary coding in mind? I love the idea, and I think you could make a killing if you are successful. I have a Master's in applied economics, and I'd love to help if I can find a way to be useful. Contact me.

My primary course of study is inferential statistics, but I have pursued minor education in microeconomics as well as other things. I became interested in utilizing econometrics for a venture such as this through private study of time-series analysis and frequency domains.

The codework will be done in-house (or will be open source) unless a situation arises where we need to use commercial code. We can judge this as the product goes through development. Thanks for the support, I'll let you know if I can use you. Probably won't be long. :)

Would be interested in using it. Keep me in the loop!

One thing I do suggest: dumb the idea down about 3 orders of magnitude if/when you go to pitch this to anyone.

daraddishman: noted. It's technical right now because I'm looking for very specific help and feedback at this point. It'll be touched up before anybody aside from development sees it, and pitched differently to investors and differently still to potential customers.

Yeah, let me know if you need some marketoid babble. I can spew the marketing gobble-de-goop like mad. I quite like this idea. :)

On the surface this sounds like an excellent idea. If initial set up and structure is sound then growth would be automatic.

I think http://www.compete.com does some of this as well but i am not sure to what measure.

I have registered only today and this is the first idea I'm visiting, I'm impressed with the idea presentation and the comments.

Thanks for all the support, guys. The project has entered development, so if you have mad skillz and think you could contribute, let me know. Code monkeys

are really what we need at this point, for I am a lone ranger and can only look at lines of syntax for so long before breaking down into fits of sobbing and wanting my mother. It's true, and I'm not ashamed. :o

Superlulz for hitting the button on accident.

Why not use the existing keyword density tools free online? What is the difference between them and your service?

Keyword density measures how often keywords occur on a given page. Keyword gravity measures, over time, how often keywords occur in relation to each other over a given number of websites that are related by niche. This service creates inferential statistical data based on this information and allows customers to make very educated guesses while marketing their product or trying to improve their site's listings. For example, one could set this software to track keyword density for two or more sites over a period of one month. We'll call these fictional websites super.com, awesome.com, and gnarly.com. All of these websites are in the surfboard business and each site's main purpose is to sell as many surfboards as possible to as many customers as possible. Each site markets new merchandise through their website as soon as it enters their inventory. This seems like it would be an obvious scenario, but for the sake of science, hear me out.

Let's say that Super is located in Los Angeles, Awesome in Honolulu and Gnarly in Cocoa Beach, a popular surfing location in Florida. Each site markets surfboards that are both internationally available and locally available. The owner of Super, knowing that his shop is in a place where trends catch on quickly, wants to start importing the newest boards from out of town...but not only that, he wants to check out which boards are creating the biggest buzz, so that the marketing part for the product is largely done for him. He utilizes this service to track how keywords are being thrown around between Awesome and Gnarly, and notice that for the month, Gweedo 240 boards have been referenced 38 times around the site at Awesome, and 27 more times beyond that at Gnarly. Not only that, but there's been a steady rise in popularity for boards that are more aerodynamic. This shows the guy from Super that while some boards are popular, not all of them may be aerodynamic. He can cross-reference this information with densities from other sites in that niche and determine just how popular aerodynamic boards are, in what places, and can generally make an assumption why. He also pays attention to other keywords that appear with aerodynamic among sites tracked back from the sites he began with (using spidering and other methods) and determines that a particular board would likely make him a killing once imported. Not only that, but he does it before his competition, simply by reading how that unique data was evolving within a month among sites that are generally trying to get the same thing done as he is. If one of those sites is above him in search listings, he could also use this information to see precisely how their SEO is done differently from his, not in regards to simple traffic, but to how those changes are affecting their rankings over time.

I think that this wealth of information is greatly underutilized (if utilized at all) and could lead to a multitude of applications, each lucrative and practical.

i lke to borrow Laura's line..this is interesitng.. brand retention is always a goal!