Workflow, automation, and control
I’ve been in the market the last few days for a web-based personal financial management tool. Managing bill dates, receipts, and so forth is an exercise in juggling and memory, and I thought a web-based solution might be optimal - my wife and I can both log in and update the accounts with our receipts, see where we are, how we’re doing compared to our budget, and all the other little things one likes to do.
It has been an exercise in frustration, though. There are four tools I’ve looked at: ClearCheckbook, Mint, MySpendingPlan, and Wesabe. Of the four, only ClearCheckbook is a viable solution, and it is very primitive compared to the others.
This isn’t really germane to librarianship, as such, but here’s my superquick review of the systems:
- ClearCheckbook: simple, clean, easy to use. Not particularly sophisticated. You create your accounts manually, input transactions manually. It does not connect to any banks or online services. You are required not to use any sensitive information - you call your checking account “Checking Account,” not “America Bank Account #123456789-0.” It supports expense categorization, simple budgeting, and reporting. Terms of use consist of 3 promises - they won’t sell your email (the only piece of personal info they have), and you in return won’t put any sensitive account data on their site, and you’ll use a good password. Otherwise, you’re on your own.
- Mint: pretty, slick looking. Syncs with your bank accounts. In fact, you cannot create an account manually - you must let them sync to your bank accounts automatically. You also cannot add transactions manually - the only thing it does is pull down transactions that have cleared your bank and put them in. This means that, for instance, you can’t add a check you just wrote; you have to wait for it to clear before it shows up. Not terribly useful, in my mind.
- MySpendingPlan: would have been good, if not for the fact that it is absolutely littered with ads, offers, and advertising spam. Has a 7-step account setup thing (with interview questions like “when was the last time you evaluated your insurance needs?”) I wanted to use it but the sheer number of “free insurance quotes” and “looking to buy a home?” and vendor-linked shopping lists sent me on my way. It’s sponsored and paid for by the American Home Buying Association or something, so…
- Wesabe: pretty, and has the interesting feature of building a tag-based aggregation of financial transactions (that is, you can tag a transaction with your own tags, and the amount of the transaction is aggregated with other similarly-tagged transactions from other users, which can let you know “is this amount normal” or “am I paying more than I should?”). Requires you to connect to your online banking services to upload accounts. Uploading doesn’t work - it’s clunky and broken. What’s more, there is once again no provision to enter transactions on your own - you have to wait until they’re downloaded from your bank. Not very useful. They do permit “cash” accounts to be manually created, though…I may go back and try that, because the aggregation features seemed nice.
So, enough about the tools - of the four, I choose ClearCheckbook, because banking info is sensitive, and I am in absolute control about what information I provide to the service. While I’m putting in transactions and such, I’m not required to put in passwords, expose account numbers, or anything like that. What’s more, it lets me put in transactions myself (like, say, transactions I just made). I don’t get the bizarre adherence to automation in that respect, to the extent that you are prevented from entering your own transactions.
And that made me consider an aspect of the “negative-click” repository idea that has been discussed previously. The goal was (to some extent) make repository contributions as effort-free as possible, by integrating all sorts of automated communication between various elements of the repository. However, there is likely a subset of users - or use cases - where the people making deposits to the repository want to have absolute control of what is added, how it is added, and how it is used.
For example, if we look at it from a scholarly workbench point of view, we’re encouraging the use of the repository along the entire lifecycle of a piece of scholarly literature, including the sensitive early part where the scholar probably doesn’t want to show his or her hand about what research he or she is performing (for whatever reason: it’s sensitive, they don’t want to be scooped, whatever). For example, I’d love a tool that supported the process, but I have research papers in the works that I don’t really want to publicize quite yet.
I see an (admittedly somewhat weak) analogy between this situation and the situation imposed by the financial management software mentioned above - if I want to add potentially sensitive or confidential data to a repository, I want to control absolutely what is added - I may not want it automatically tagged, uploaded, and connected to other similar research. I may want to simply have the data there, but avoid any really specific information associated with it - basically, make it something that only has meaning to me.
This tends to fly in the face of some of the open-access goals of repositories, it seems to me. But if there is a goal to create a workflow-integrated repository with minimal effort, it strikes me as something that should be addressed.
So, gentle readers - if I have any gentle readers of this post - what are your thoughts? Is this an interesting problem? Is it irrelevant to repositories? Is it simply so obvious as to not bear mentioning?
You may be interested in Monash University’s work, which explicitly posits a two-layer repository, one for in-progress work and one for later dissemination and preservation.
This isn’t just a problem, it’s THE problem when talking to many antsy faculty. They would love to have storage for in-progress work, but they’re hyper-competitive and hyper-paranoid. And if you don’t give them storage for the in-progress work (which is when they actually perceive they need it), they don’t care about storage for the final result, which is obviously a problem for an IR!
Dorothea, thanks - I’ll go take a look at that now, in fact. I’m in charge of - well, they call me “coordinator” - a digital repository project that’s just getting underway, and one of my biggest concerns is selling the thing to our faculty.
Weeeellllll… here’s the thing.
You can’t. There isn’t enough there to sell, with most repositories.
What you need to be able to sell are services, keeping in mind that “preservation” and “increased citations” are just not services that faculty have been shown to value. The best advice I have is to figure out what services you want to offer that faculty would bite on, figure out the gaps between those services and what you actually can offer, plug them, and then go forth and market.
This is hard. A lot of libraries got into this business thinking that repos would more or less run and fill themselves without much intervention. That didn’t pan out, so some of us are going around for Round 2. For a good example of a Round 2 repository, look at Ohio State.
I used the word “selling” a bit too broadly - I was really getting at “what part of the package will people really want,” but I take your point. That is the worry we have at the moment, identifying the services that will be valuable.