Web Finance, Past and Future

Shortly after I started my first job, and started getting paychecks and managing actual expenses, I created my first personal finance spreadsheet. I wanted a simple way of tracking how I spent, when and what, essentially, and a spreadsheet afforded the most straightforward way of doing this that I knew.

The spreadsheet was never a regular focus of my finances, but something I would update every few months or so, populating the data with electronic bank statement downloads or by plain old data entry. It was a pain in the ass, but I didn’t need to do it frequently.

I even gave some dedicated desktop programs a try, including Quicken and Microsoft Money. While they were definitely cooler than my spreadsheet, they were overkill and didn’t really do much for me other than offer a password for the data.

Enter Wesabe

So I was excited when I discovered Wesabe. Wesabe was (the financial management side of the site shut down a few years ago) a personal finance site with a focus on community and a flexible money management system based on tagging. It was easy to track spending down to the penny and the social aspect of the site helped engage people and created a friendly face for the service.

Updating my Wesabe data was a matter of entering transactions (for cash based purchases, which were rare for Mr. Post-cash Economy dude) or uploading a bank transaction file. The thing I liked most at the time was Wesabe’s completely customizable tagging interface for tracking expenditures, rather than categories. This provided a lot of flexibility and make updating transactions fast.

Of course, updating transactions - even when it takes only a few minutes - gets to be a royal pain in the tuchus. So I gave Mint a try.

Have a Mint

Mint was somewhat similar to Wesabe, but the biggest differentiator was that Mint scraped all your data for you. And its category recognition was pretty good, based either on the merchant data in the transaction or your own pattern of categorization. Switching from tags to a hierarchal category system was a bit of a drag, but the categories were easy enough to edit, and Mint got rid of the clunk category drop down in favor of a suggested search text field. Now it didn’t matter where in the category tree your transaction existed.

I did have some more serious concerns adopting Mint, and those finally contributed to why I deleted my account altogether.

As you can guess, if you want Mint to automatically scrape and update your data, it needs access to your online bank accounts.

At the time I was trying it out I had also read some posts (probably on Hacker News) about the security at Mint, and came away feeling relatively confident that my access credentials were secure. Any worries I had later were smoothed over by the slick convenience of the system.

Then a couple things happened. Mint was sold to Intuit and my primary bank updated its site. The former meant nothing specific to me as a user. However, even as a happy TurboTax online user, something didn’t sit right with that relationship. No concrete reason to which I could point. And my bank’s updated security meant that Mint was increasingly asking me the answers to my bank’s secret questions, which I’d never provided as something of a last ditch attempt at maintaining control.

Two things finally convinced me to delete my account. One, I wasn’t making decisions based on my use of the site any longer. And two, I decided I needed to winnow down the number of services I use in which I’m the product. Consumers aren’t really part of the information economy any more so much as they part of the personal information economy, and I realized I didn’t want that kind of information (credentials and financial history) in the hands of someone for whom I was not the customer.

Proprietary bank solutions

It’s worth mentioning, if briefly, that some banks offer similiar personal finance features through their consumer web sites. Bank of America and USAA both allow you to link third-party bank accounts and provide aggregate transaction data.

I don’t remember ever giving the Bank of America service a try, and likely passed it over without a consideration given the terrible interface they’d provided elsewhere. USAA had a pretty good web interface and compared to other banks I have a high level of trust in USAA. So it was worth giving it a go, but turns out that personal finance management software - interfaces, especially - is something demanding a pretty deep commitment.

The USAA personal finance management interface consists of a set of rigid categories and largely meaningless pie charts. While USAA is good at banking, turns out they’re not that good at interface design.

Hello, HelloWallet

The latest service I’ve tried is HelloWallet, which bills itself less as a straight money tracker and more as a financial guidance service. It’s got two big things going for it, so far as I’m concerned. One, it’s proudly made in DC, and two, it’s a paid service. Yes, that’s a benefit. Since I’m the one paying them, I’m the one to whom they have responsibility. There are no affiliates or advertisers in the mix.

On paper I love HelloWallet. The customer incentives are in the right place, the service is focused on helping customers achieve financial goals, and they actually have people driven customer service.

But the lifeblood of any financial management system is still transaction data. Budgets and goals are meaningless if you don’t know how you’re tracking, and you don’t know how you’re tracking unless your transactions are accurate, up to date, and propertly categorized. It’s the last part where HelloWallet made me pull my hair.

Identifying the nature of a transaction from bank data is really hard, and so predictive categorization is only so accurate. This is to be expected, and it’s one reason why every piece of personal finance software I’ve ever seen lets you edit the transaction category. The other reason they let you do this is that everyone has their own way of thinking about and measuring their finances. Spending at coffee shops may be an entertainment expense to one person and a food expense to another. I’ve always thought of coffee expenses in the latter category, while HelloWallet sticks them in the former.

Both the built-in category structure and the method for updating transactions in HelloWallet feel broken to me, or at least very incomplete. Editing transaction categories is really hard, or at least significantly harder than it needs to be.

The built-in categories feel a bit off and overly rigid. I’ve already mentioned the “Coffee shops” categorization. Then there’s the top-level “Children” category. If most customers have children then a default top level category for children might make sense, but it’d also make sense to let everyone else remove that category. You can’t edit or remove built-in categories. As it were, I think of children’s clothes as a subcategory of clothing, although that makes it harder to later show your children in aggregate just how much they’ve cost you!

Every application and service comes with minor annoyances to at least some users. In the case of updating transaction categories it’s a matter of usability, not just for the interface but for the application as a whole. I really want to like HelloWallet again, and these problems are solvable, so I’m hopeful that the team there will address them. I’d happily come back.

Personal finance software tomorrow

From the systems I’ve used so far, something mixing the HelloWallet business model with Mint’s data interface would come closest to the ideal.

Because the ideal system is one in which keeping your transaction history up to date and accurate is frictionless, the system adapts to your personal financial model, and it stays one step ahead of you, keeping you out of trouble and on track to your goals.

Maybe that’s Simple (neé BankSimple), although not having had the opportunity to use it yet, I’m not really sure how it fits into your personal finances in conjunction with third party accounts.

Originally published November 2011

Are you a Django developer? Have you found yourself wishing you could reuse the code you write or looking for some guidance on building standalone Django apps?

Check out my upcoming book, Django Standalone Apps: A developer's fieldguide to developing reusable Django applications the first guide written specifically for writing standalone apps.