Service Review: Wesabe's open source finance tool

Years ago, my wife and I used to religiously enter everything from our checkbook into Quicken. Unfortunately, we did very little analysis of how we were spending beyond "too much" or "just right." We knew exactly how much we were spending, but not why or where.

Years have passed, and we have become even worse about managing our money. When big or out-of-the-ordinary chunks of cash (bonus, consulting, whatever) come in, we're good at applying that to car loans/etc. such that we have no debt beyond our mortgage. But we still stink at managing our money, in part because we don't have anyone advising us on how smart people manage their money.

So, today, I gave Wesabe a spin. I've known about Wesabe well before it became a company, having discussed it with Marc Hedlund while he was still an EIR with O'Reilly Media. The basic idea: harness the power of a community to analyze one's spending and to get collective help (tips, etc.).

Open source finance, if you will.

Wesabe is not, of course, open source in the traditional, OSI sense. It's about open collaboration (Web 2.0), but also about open data.

Does it live up to its billing? Mostly. I found the analysis tools to be wanting, though even just having a tag cloud was an excellent way to see where the Asays spend their money (the answer, incidentally, is "food," "household," and "kids" - I'm sorely disappointed Arsenal doesn't get more of my money :-). It would be great to have the tag cloud represented in tree map form - I've become a fan of these, and think it would be a better way of representing where one spends money, and whether that amount is trending up or down (especially if measured against one's goals) on a monthly basis.

Wesabe also does a poor job of recognizing vendors. My bank represents vendors like this:

24323016QG9VHN3Q0 DOMINOS PIZZA #7523Q09 SALT LAKE CITY UT

That's Domino's Pizza (jalapeno and pepperoni, thank you very much). Unfortunately, sometimes Domino's Pizza looks like this:

DOMINO'S PIZZA 07523 SALT LAKE CITUT US

Wesabe's system didn't seem to recognize the common theme between these (Domino's Pizza) and made me enter it in each time....Painful. Lots of times. We really like Domino's.

It was the same for Gymboree, Michael's Crafts, Chevron, etc. This seems like a very easy problem to solve. Perhaps Wesabe is concerned about false positives, but I'd rather have a few errors than having to re-enter all of my data after importing it from my bank. To be fair, it did well with Dan's Foods (grocery chain) and others. But it had maybe a 50% success rate.

That's the bad news. The good news is that it's a young company with a robust community that will only keep getting better. Each correction I made to the system, for example, in labeling my transactions fed back into others' transaction labels. Every tag I added (which I did simply so that I could personally track my own purchases) improves the system for everyone else. I started reading tips that others had provided and plan to look into a few of them more closely.

So, it's a system with warts, but enormous potential. I'm now a fan, despite spending nearly two hours fixing four months of checking account data. Here are some things that I believe will make it a better system:

1. Suggest tags for a vendor/purchase based on how others have tagged it. Today, the system expects the user to do this work.


2. Correct obscure bank nomenclature for vendors/transactions based on the collective intelligence and (better) keyword analysis.


3. Let me know how my spending in given categories compares to others'. I'd like benchmarks. You can drill down in categories to find out how other Wesabeans spend on that category, but I want a global view on spending and I want to compare like to like.
   

   For example, he system tells me that other Wesabeans spend $210/month on food. These people clearly do not have four kids like I do, and it doesn't tell me if they define "food" as "anything that comes from a grocery store," as I do because I don't have a more granular way to do it. For "groceries" Wesabeans spend $177/month. I'm lucky to get out of the grocery store in one visit at less than $177. Wesabeans must be single. I don't want to be compared to them and I don't care much for their advice. I want to hear from others like me.



4. Reach out beyond the Bay Area. The system needs to represent a diverse range. Today, based on the tips and the spending priorities, it clearly does not.


5. Align appropriate tips with users at the right moment in the system. Don't make me go click for tips. If the system sees that I'm spending a lot on gas, as I hover over that bar in my spending analysis, tell me what other people with kids suggest (and do not tell me what some Berkeley-ite thinks about eco-friendly cars, because I don't care).


6. Augment the tag cloud with a tree map view.

These are just a few suggestions for a useful tool. Again, the tool is useful as it already is because it gives my wife and I a common view into our finances. We don't have to hover over the other's laptop. I'm a fan, and will become even more so as the system draws in a more diverse crowd and then applies that collective intelligence in, well, more intelligent ways. Good work, Marc and team. The good news is that there's still lots of work to do.