How to most efficiently reorganize a filing system?
July 26, 2006 12:18 PM Subscribe
I am looking for a resource or advice on organizing files and folders on a shared hard drive. I am faced with 12 GB of files that need to be reorganized - many files and folders are out-of-date, irrelevant, or obsolete, and many were created by people who no longer work at my company for projects that are no longer active. I'm hoping to hear about a strategy that helps me methodically organize all of this data, archiving what's no longer needed and coming up with an intuitive filing system for what is.
My methods so far have been too haphazard - for example, I'll go through every Excel file or every PDF in a subdirectory, then move on to another file type. I have ended up with folders called "archive" in every directory, some with a couple files and some with hundreds. There are lots of different versions of documents saved in different formats, filenames, folders, with no consistency. Followers of 43folders/GTD/lifehacks... I'm looking for strategies but I'll accept tactics too!
My methods so far have been too haphazard - for example, I'll go through every Excel file or every PDF in a subdirectory, then move on to another file type. I have ended up with folders called "archive" in every directory, some with a couple files and some with hundreds. There are lots of different versions of documents saved in different formats, filenames, folders, with no consistency. Followers of 43folders/GTD/lifehacks... I'm looking for strategies but I'll accept tactics too!
How big is your organization? How is your paper filing organized? Do you have records retention schedules for your paper filing? Do you have administrative professionals in the organization (IE: experienced secretaries, filing clerks, etc)? What sort of work does your office do?
posted by raedyn at 12:41 PM on July 26, 2006
posted by raedyn at 12:41 PM on July 26, 2006
Response by poster: (baylink - yes I have seen other AskMe threads about personal file organization - however I am in the strange situation of organizing employee Q's reports in a way that salesperson P can easily find them!)
Background: We are a very small web design company. We have almost no paper filing, but what we do have is also haphazardly organized with no record retention policies. I'm the closest thing to an administrative professional we've got, but I'm far from an experienced secretary.
posted by pants at 12:50 PM on July 26, 2006
Background: We are a very small web design company. We have almost no paper filing, but what we do have is also haphazardly organized with no record retention policies. I'm the closest thing to an administrative professional we've got, but I'm far from an experienced secretary.
posted by pants at 12:50 PM on July 26, 2006
You've really got two three problems: Coming up with a workable system now, applying it to your archives going backwards, and enforcing going forwards.
To answer #1, my own hierarchy goes ClientCode/Year/JobCode/[job files]. Pretty basic. It almost doesn't matter what scheme you use, as long as you use it consistently and it reflects the information you need.
To answer #2, there's no easy shortcut. You'll need to look at everything and apply your system to it.
My wife works in a graphic-design shop, and deals with #3 constantly. They've got stuff live on a server and some stuff archived to DVD-R. They've got a formal file hierarchy and nomenclature system (more complicated than mine). All active job files are supposed to be checked off/onto the server. And yet, some of the employees keep files on their desktop, and name them whatever they feel like. If you come up with a way to prevent this, I'm sure my wife would love to hear about it.
posted by adamrice at 1:11 PM on July 26, 2006
To answer #1, my own hierarchy goes ClientCode/Year/JobCode/[job files]. Pretty basic. It almost doesn't matter what scheme you use, as long as you use it consistently and it reflects the information you need.
To answer #2, there's no easy shortcut. You'll need to look at everything and apply your system to it.
My wife works in a graphic-design shop, and deals with #3 constantly. They've got stuff live on a server and some stuff archived to DVD-R. They've got a formal file hierarchy and nomenclature system (more complicated than mine). All active job files are supposed to be checked off/onto the server. And yet, some of the employees keep files on their desktop, and name them whatever they feel like. If you come up with a way to prevent this, I'm sure my wife would love to hear about it.
posted by adamrice at 1:11 PM on July 26, 2006
I believe this MeFite's question is comparable to yours. Hopefully, you'll find the answers it prompted to be valuable.
posted by NYCinephile at 1:18 PM on July 26, 2006
posted by NYCinephile at 1:18 PM on July 26, 2006
pants, the task basically falls into two phases.
Phase 1 is where you find out from P what ways P would like to access Q's reports, and from Q what kind of meta-information Q used to label and file the reports. P's requests will probably not conveniently intersect Q's labels or methods, so as the output of Phase 1 you build and get agreed a cooperative taxonomy to guide your Phase 2. This need not be terribly onerous, but getting agreement by all parties is vital. P may need info by customer, by date, by cost, by project type, and by various compound or concatenates, i.e.. projects by customer/project type in 1999, or customers/project cost (biggest revenue sources).
Once you've got your taxonomy agreed, you construct a filing and retrieval system to reflect that taxonomy. If both P and Q natively "think" in terms of customers, and Q is labeling his reports consistently by customer, then customers is a primary organizational structure in both your filing and retrieval mechanisms. In your primary org structures, such as customer folders, you need to group and segregate your documents by the most broadly common secondary structure. Customer folders might have cost, code, content and correspondence secondary sub-folders, or whatever best supports the concatenates of your taxonomy. You finish Phase 2 by creating and publishing reasonable retrieval indexes to support and extend the use of your taxonomy for ad hoc requests.
posted by paulsc at 1:25 PM on July 26, 2006 [1 favorite]
Phase 1 is where you find out from P what ways P would like to access Q's reports, and from Q what kind of meta-information Q used to label and file the reports. P's requests will probably not conveniently intersect Q's labels or methods, so as the output of Phase 1 you build and get agreed a cooperative taxonomy to guide your Phase 2. This need not be terribly onerous, but getting agreement by all parties is vital. P may need info by customer, by date, by cost, by project type, and by various compound or concatenates, i.e.. projects by customer/project type in 1999, or customers/project cost (biggest revenue sources).
Once you've got your taxonomy agreed, you construct a filing and retrieval system to reflect that taxonomy. If both P and Q natively "think" in terms of customers, and Q is labeling his reports consistently by customer, then customers is a primary organizational structure in both your filing and retrieval mechanisms. In your primary org structures, such as customer folders, you need to group and segregate your documents by the most broadly common secondary structure. Customer folders might have cost, code, content and correspondence secondary sub-folders, or whatever best supports the concatenates of your taxonomy. You finish Phase 2 by creating and publishing reasonable retrieval indexes to support and extend the use of your taxonomy for ad hoc requests.
posted by paulsc at 1:25 PM on July 26, 2006 [1 favorite]
Best answer: I did something similar for the shared-by-everyone folder on our library LAN. First thing to do is decide what you'd like the new top-level directory to look like and create a new structure of empty folders to move files into as you organize; that's the easiest way to tell where you've been. Keep the number of top-level folders minimal and make them mutually exclusive.
Then, create a folder for "dead" stuff. Pick a date; everything older goes in that folder, without exception, retaining its current folder structure. This saves you having to ask if people are using old files created by people no longer around. If they complain that they can't find a file, you know it's in the dead folder. (This time next year, anything still in that folder can likely be dumped.)
Once the old stuff is out of the way, sort the remaining documents and folders into your top-level folders, and explain the system to everyone. We don't allow anyone to make new top-level folders, so at least that minimal order is preserved.
posted by donnagirl at 1:52 PM on July 26, 2006
Then, create a folder for "dead" stuff. Pick a date; everything older goes in that folder, without exception, retaining its current folder structure. This saves you having to ask if people are using old files created by people no longer around. If they complain that they can't find a file, you know it's in the dead folder. (This time next year, anything still in that folder can likely be dumped.)
Once the old stuff is out of the way, sort the remaining documents and folders into your top-level folders, and explain the system to everyone. We don't allow anyone to make new top-level folders, so at least that minimal order is preserved.
posted by donnagirl at 1:52 PM on July 26, 2006
This thread is closed to new comments.
posted by AllesKlar at 12:25 PM on July 26, 2006