…even if a boss decrees you will use it, you have control over the end result.

This does seem to be an area that people find interesting. Quite a few people have said to me that they had considered doing a similar thing. There are other competitors out there, some of them reasonably pricey (not necessarily a bad thing).

I was also uncomfortable about the monitoring thing, and essentially make it easy to edit/change the data if you want to. You can add or delete entries, pause the monitor, and specify an activity instead of having it automatically recorded. So even if a boss decrees that you will use

I am afraid that exists too, so don’t beat yourself up too much about it.

Huh. I hadn’t even though of that. But you make an excellent point. I’d have to give some consideration to making the software really useful for individuals while minimizing the spyware capabilities that might be of interest in the employer/employee monitoring niche. Becuase that’s not a niche I’m particularly interested in exploring with my software company.

But I see a much darker use in the distance, by companies wanting to monitor what their employees are doing all day. Yeah, I know they have a right to do it and all that, but it would make me very uncomfortable, regardless.

Maybe you could create a tool to trick the program into thinking you were doing something else, then you could sell both and hedge your bets. If you kept the updates at around a 6 month offset from each other, you could milk that cow forever……

*sinks into deep thought*

Actually, the software doesn’t try to form any conclusion about individual companies, or individual securities. Instead, it just tries to find consistent patterns in the marketplace.

For example, here are the kinds of reports that the software can produce:

When a $5 ~ $10 security decreases in price by at least 3% each day for five consecutive days, while trading volume increases by at least 5% each day, then the stock is most likely (68% probability) to lose 13% of its value over the next ten trading days. It is next most likely (24% probability) to level off and maintain its value over the next ten days. And, in 8% of cases, the security increases in price by 1% to 5% over the next ten days. This pattern has been detected 1689 times during the last ten years, so it has a standard margin of 2.43%.

Something like that.

The idea is not to predict the movements of particular securities, but to identify reliable, repeatable patterns in the market.

And maybe today the pattern is detected on MSFT or CAT, but next week it might apply to IP or JPM.

Of course, those patterns change over time, so I’d give more credence to a pattern that occurred commonly over the last year than to a pattern that occurred commonly ten years ago.

Idea #20: Desktop Analytics
I remember seeing someone on BoS who mentioned doing this already*. It was a tracker with a 1-minute granularity, and it would look at the window in focus, etc.. If you search in there, I am sure you’ll find it.

Idea #24: Product Info Search
Again, a bunch of different people are already doing this*. Here are two:
http://www.customizegoogle.com/
http://bettersearch.g-blog.net/

Idea #25: MicroISV GoldRush
I like this idea.. Especially if you have physical access to where these people hang out. Internet friends are nice, but people still favor the good’ol face-to-face interactions.

Idea #27: Lick This Chocolate
This is one of my most favorite of all of your ideas. I can just see this being a huge hit on Valentine’s day, birthdays, christmas, haloween, etc.. If you can do it, go for it! If you could get a FoodTV network person to incorporate it in their show, you got a HUGE winner! I can see Rachel Ray giggling, moving her hands around and showing this chocolate off!

Idea #28: VaporTrade
I can see your background popping up in this one. It is very neat. Reminds me of Mr.Data shuffling through the Federation database, looking for patterns to uncover a conspiracy. However, I don’t know how useful it is to look for patterns like this for applications in the stock market. This would work nicely, if you got a lot of constants like same people running the company for a long time, keeping the same overall strategy, etc.. Most large companies probably fall into this category, but other smaller, unknown ones might not. And your pattern search might be comparing apples to oranges giving you misleading information. Then again, you can always put a disclaimer.

*When I say “someone is already doing this”, it is just to give you a direction and not necessarily to say “This is a crowded market, avoid it”.

Or more generally, what you focus on expands.