Line of regression.. need some guidance


I was wondering if you could guide me a little so that I can head off and do some analysis..

I'm looking to calculate
1. the time it takes for my team to close a business deal and
2. what our closing rate is. i.e rate of success

The information I will need is :
1 - the start date of all my previous deals and
2 - the close date of the deals.
To get my closing rate do I need to create a separate column to put the closing dates of deals we have closed?

Once I have the successful closing dates in days I can calculate the time it takes to close a deal.

Can I get both information in one go or would I have to do two different regressions?

In the formula y = a + bx + e
Where y = time to close a deal?
what is a?
How do I get the error i.e. e ?
Would b be the slope i.e my closing success rate?

How do I get the line of regression so that it can help me forecast how long we should spend on a deal and then give up?

Any tips?
Your guidance will be very helpful

Thanks a lot..

- Lost -


TS Contributor
Unfortunately it looks like you posted during a time when we were away from the forum - summer vacation, etc., and when we came back we just started responding to more recent posts.

I'll take a look and get back to you.


TS Contributor
From what you are describing, you'll be able to compute the average time it takes to close a deal, and the standard deviation of that time. You'll also be able to determine success rate (# deals closed / total number of deals).

But that's about it. In terms of regression, I don't see how this data will enable you to "predict" how long you should let a deal go before giving up on it.
here are some thought on this,

you only have one variable, time to close, so there is no opportunity to do a regression. unless you want to try to plot time versus cumulative % closed?

it would probably not be a linear function though,

you could likely find the point at which 95% of all deals have closed and use that as a cut off point.



TS Contributor
This assumes, of course, that "time to close" and the success/no success variables are related to begin with....