Earlier this week, my boss and I were reviewing on the status of our automation project. We were really pleased at how everybody was coming along on the report, even figuring out what to do when errors popped up. It’s such a good feeling to see it come along so well. There were just a few lingering issues that my boss and I were going over, mainly one particular type of entry that often causes imbalances – something called waitlist. After some research, I found that none of the required reports showed any information regarding waitlist so I offered up a solution where folks could enter the number of waitlist on the file and the file would incorporate it appropriately into the statistics. This suggestion led to an even better idea from my boss: provide a step by step checklist on what to do when a certain error popped up.
The concept is a cool concept: when an error message pops up, step 1 of the checklist appears and asks you a question. You answer the question in the yellow box. Depending on the answer you provide, instruction tells you what to do and step 2 pops up. You continue and follow along in such fashion, answering all of the steps until there are no more steps. The checklist even had 3 scenarios for the waitlist problem that we were encountering, and once the answer to the waitlist questions were answered, the spreadsheet handled the waitlist issue. Now, myself, I never had a need for such checklist so I had never developed one. Instead I developed automated files. But when designing a file for others to use, this was a cool concept and hopefully, a good one. We’ll see this Monday.
The other idea I had, or used, was the idea of using data tables as a simulation method. This idea was something I learned last week when I was working on the Sunday Chronicles’s Simulation project. Normally, I don’t use data tables but I found an opportunity this week to try it out as a way of building a simulation. I’m working on a way of figuring out potential revenue based upon the idea that so many leases expire in a month and only so many folks renew their leases. We have at a macro level a rough percentage of who will renew but we don’t know the specifics of who. Depending on who renews, you could have varying results for potential revenue. Right now I’m using a random generator to randomly assign who renews. I did it for one month and now I’m in the process of figuring out how to use it for multiple months. Once I figure out how to generate a random selection of who renews each month, keeping in mind the overall renewal percentage, then I want to build a data table of multiple simulations. By making multiple simulations, I hope to use it to see if the potential revenue tends to home in to a certain value. Does the set up narrows to a small band of revenue potential or does the revenue potential oscillate wildly forever?
I’m not done yet and I may not succeed, but I sure hope I do. It is an interesting way of doing simulations and I would like to see the outcome of doing such a simulation.