All in on Devonthink 3 and couldn't be more pleased

Like another user on here, in the past several months I had to declare organizational bankruptcy. I am a solo criminal defense attorney. And I receive a lot of paperwork. The amount of paperwork per case can range anywhere from 10-20 PDF files to hundreds of PDF files. A plus is that now that most of the Courts I appear in are in e-filing, I don’t have to do nearly as much scanning as I used to. Nevertheless, I just couldn’t find the time to do the task of moving all those PDFs to their assigned folders. I gave up. I had an inbox with 750 PDFs in it and would occasionally spend 30 minutes looking for a single document that I needed for a hearing.

So I decided I would hire a virtual assistant to help me organize my digital mess. I even started preparing some instructional screencasts for whomever I would end up hiring. As a last ditch effort, I decided to index my 70 GB of work folders to see how fast it would process. It seemed to work well, so I decided to go ahead and import a copy of the whole thing. In one day, I easily processed all my PDFs using the classify shortcut (ctrl-c). I have found that Devonthink works very well at classifying documents if you have flat folders with no or very few subfolders. The more subfolders, the more DT seems to get confused about where you want to put your files. I basically just have flat folders organized by client name (last, first). Instead of using a lot of subfolders, I have created about 25 tags to apply to the individual files. DT3 now has a filter pane that allows you to filter files within a group. This basically operates with the same effect as subfolders would. Although tagging is time consuming, I am trying to limit both tagging and naming conventions to cases where there are a large number of documents, such that I can’t keep track without those.

I had previously tried to use DT2 with my old computer and, I am assuming due to lack of available memory on that machine, the database would slow to a crawl. But with my new MBP I have found DT3 to process the information very quickly. Still, I was wary about having a mega-database, which would eventually become bogged down. But with DT3, classification now works across all open databases. Currently, I have a database with only the folders of cases I’m actively working on. This database is only about 20 GB and I have it synced with DTTG. The other 50 GB of closed cases is in my archive database. Of course, if I had an extremely complex legal case with thousands of PDFs, then I would create a stand alone database for it. I have not noticed any problems with the classification working across databases.

I am now refining my workflow regarding annotation tools and analyzing documents in Devonthink (love that I can basically use my DT database with any application that supports hyperlinks, so that I can create outlines and chronologies with a link to my documents in DT3). But just getting my files organized has been a big weight off my shoulders. I actually feel I can find what I’m looking for. And the classification has worked great for me. On one occasion, I used auto-classify on about 500 PDFs, and I didn’t notice any problems. I know that my situation might be drastically different from a lot of people in this forum, but just felt like sharing my solution to a problems that has vexed me for the past 3 years. Now onto the next problem!


If your documents and tags follow any kind of pattern, you might be able to use Smart Rules to auto-tag the things you’re dropping in. Manually tagging things is something I abhor!

You might appreciate the script discussed in this thread, where a couple of users have figured out how to save links to precise lines in a given document.

Also, make sure you share your story with the DT crew. I think they appreciate these tales!


Good point. I don’t disagree. Manually tagging (and filing) will probably result in more meaningful organization, as there’s abstract and abductive things your brain can do that the computer simply can’t.

That said, tags—even programmatically generated ones—offer a couple of things that are above and beyond the queries you describe. One is browsing. It’s cognitively easier to look at a set of tags as cues as you’re looking for something and to click one than to generate a query for the same set of features represented by that tag. Instead of a tag, though, this could be implemented through some kind of saved smart search.

The second isn’t really related, but worth mentioning (even though it’s obvious)—a file can’t really be in multiple subfolders (with exceptions, e.g., aliases in Finder and replicants in DT) but it can have multiple tags.

But yes, in general, manually organizing stuff is probably better if you’re willing to put in the effort. (I am not. :upside_down_face:)


I presume getting clients to pay on time? Don’t know if Devon can help there, I jest of course I feel your pain my mother went Solo and faced the same avalanche of documents even after eFiling. If you don’t mind keeping this topic periodically updated that would be excellent.