## Help with average length of time over time

Hi all,

I'm hoping someone here can help. I am trying to calculate a number which is the average length of time a customer is with us. The service we offer is a monthly service (similar to cellular phones minus the contracts). I'm having some difficulty with how to accurately calculate this.

To me there are three ways to try to get this without a more powerful formula. I will propose all 3 and then give my reasoning why none of them seem to be truly accurate. What I am in need of is the more powerful formula

1) We could take the average of all people that have been with us and since cancelled service – but then we don’t take into account the numerous clients that are actually with us still.

2) We could take the average of all clients that have been with us and cancelled (above number) AND clients that have not cancelled yet but we make today their cancel date – but then the people that have not cancelled yet presumably will stay with us longer and hence bring the average out (for example all the new clients that have been with us only 1 or 2 months will skew data to the left, since our sales have not flattened out in our history)

3) We could take the above group (#2) but limit it to those that have signed up more than 1 year ago (arbitrary) – but that too seems inaccurate since conditions may have changed so more new people could be cancelling today than 1 year ago.

The goal is to see if we are doing a better or worse job keeping customers with our service.

Thank you very much for your help!