I’ve been slipping into an OCD spiral with regard to our finances lately. This is probably 10% a reaction to the state of the economy and 90% a reaction to having a child and realizing how little money is left over at the end of the month. This was all precipitated by “The Quicken Disaster of 2008″ when I reformatted my hard drive a month ago and completely forgot to backup the last five years worth of quicken data. The last time I had backed it up was three years ago. (Hey, I’m not claiming to be a genius) I started to dive back into reformulating my financial profile and realized I had never really thought much outside of the Quicken for Mac bubble. I’d heard of sites like Mint.com and Quicken Online before, but had kind of ignored them since I already had a five year investment in Quicken. I sort of welcomed the chance to take stock. I very well could write a post titled “Why Quicken for the Mac sucks more than anything on the face of the planet”.

Thusly, this has kicked off one of the most annoying and frustrating comparison exercise of my life. To say that I’ve become more obsessive about my financial profile would be a vast understatement. I check graphs and formulate alternate budget scenarios three or four times a day. (Actually, that’s a lie. It’s more like 20, but that’s incredibly embarassing) I seriously like to visualize data. I’d done it a bit with Quicken 2007, but had always been averse the data entry piece. Receipts would pile up until I had no choice but to enter them. I’d actively encourage the use of cash for small purchases so I wasn’t entering 20 receipts for $1.20 bagels.

I started with Mint. It seemed to be the answers to my prayers. Enter my bank account info, it sucks data down, renames the ugly bank-speak to something pretty, categorizes based on community (or my) input, and presents some pretty graphs. This led me to the reverse philsophy on debit cards. Charge EVERYTHING! Put every insignificant purchase on the credit/debit card and I get an automatic and current view of where my money is. This instant gratification drove me to do more budgeting and obsessively staring at the data. So, now I care more about where my money goes than I ever have before.  A good thing, right? Yes and no. Good because I’m paying attention, bad because none of the sites can really follow through with a perfect system.

Designing a System

In addition to knowing how much money I’ve spent on haircuts in the past year, I want a fairly simple view of five general categories of expense:

  • Recurring (generally predictable costs; e.g. bills, subscriptions, mortgage, etc.)
  • Debt (money toward paying short-term debt)
  • Essential (variable costs that must be paid; e.g. groceries, fuel, dog food)
  • Discretionary (variable costs that are more optional; e.g. dining out, shopping, entertainment)
  • Savings (money put toward savings)

This allows for easy budgeting and keeping things straight. I take a short-term view when it comes to budgeting. Translate monthly targets into weekly targets, and then to daily targets. I want to visualize the weekly targets. I want to set up a simple budget for a week, and have the system e-mail me daily reports on all those targets. I want those reports to be pretty, with useful graphs. I want those reports to show how much of the spending is coming from me, how much from my wife, and how much we share responsibility. I also want my wife to get these reports and have an account where she can login and view the data. Sounds fairly simple, right?

Mint.com

Mint has some of the strongest and weakest visualization around. It’s strength is in its flash-based pie-chart graphs that let you drill through your costs like its nobody’s business. You can add subcategories (given a recent system upgrade) but not top-level categories. You can define a set of tags and apply those to individual transactions. It’s easy to edit multiple transactions at once. Mint remembers the names that you give each transaction (most transactions from the bank look like this “ANGEL AS KINKY DOLLHSE- AXGHBD0001923″) and the categories you assign it. When you fix one, it fixes all of them.

Now, onto the drawbacks….

  • The support sucks – read through the Mint forums and you’ll immediately be greated with 4.7 million posts about banks that don’t work and features that are buggy. You’ll also find about 10 responses from Mint staff.
  • No way to do pending or future transactions – all I can see is what’s been spent. There’s no way to anticipate checks or future bills
  • Can’t assign a transaction to multiple categories – This precludes me from doing both the granular categorization (haircuts) and the general categorization (discretionary) at the same time.
  • Can’t Learn Tags – Ah, you say. But there are tags. Create tags for the five general categories. Well, yeah, but you have to do them one by one. It’s too many clicks to get to the tags to make that useful. You click, wait for the dropdown to show, click on the tags, then apply.
  • Can’t budget based on tags – The budgeting system is OK in general, but it can’t budget based on tags at all. So, I can’t do what I really want: create a budget for Discretionary expenses.
  • Can’t do anything other than monthly budgets – this really sucks. All I can do is set a boundary on the first and last day of the month. That gets rid of my ability to do weekly expense budgets. But it also poses other problems. My wife and I get paid monthly, on the last day of the month. We have a lot of expenses that are due near the first of the month. If the last day of the prior month falls on a Friday or weekend, I need to pay my next months’ expenses *before* I get paid so my bank’s billpay system has enough time to get the funds to my payees. That means budgets tend to clump costs in wrong months periodically and throws off my analysis.
  • No way to get intersection of tags – If I want to visualize, say, how much money I’m spending on discretionary items versus my wife, you might think I could just create a tag for myself and one for my wife. I can, but I can’t visualize it or filter on that in any way. All I can see is all costs sourced from me, and all discretionary expenses. Lame.

Summary: Mint has some great potential, but it’s budget visualization and lack of advanced tagging features really torpedoes it for my purposes

Quicken Online

Quicken seems like such a dinosaur, but it’s interface has improved since I first looked at them months ago. Like Mint, they also learn your transaction mapping to readable names and categories. They have really good budgeting features and show you your monthly averages. It’s easier to show custom date ranges on the visualizations and reports. You can add manual transactions that haven’t made it to your bank yet. They just added a new feature on November 21st (2008) that shows a line graph of your available cash for the next two paycheck cycles (yeah!) that takes into account all your upcoming transactions. It shows very clearly how much cash you have left over when all your future obligations for the upcoming budget cycle are met. This is IMMENSELY useful since I normally just do all this in my head.

But no one is perfect…

  • No Subcategories – What? You can’t do Auto:Fuel and Auto:Repair costs. It’s all just Auto. Weird.
  • No Transaction Splits - Huh? Costs are all in one of the already crippled categories.
  • Can’t assign a transaction to multiple categories – This precludes me from doing both the granular categorization (haircuts) and the general categorization (discretionary) at the same time.
  • Pie charts aren’t as useful – They only show the top 6 categories. You can’t drill down into them. Why would you want to? There are no subcategories!
  • Budgets are monthly - Can’t choose a budget cycle on anything other than the 1st through the last day of the month.

Summary: Quicken is showing some potential. The lack of subcategories, splitting and multiple categories is still too limiting to be useful.

Wesabe

Wesabe takes a bit of getting used to. It’s entirely tag-based. You rename and tag each transactions. Tags are categories and categories are tags. But tags can be more. I could tag something as “Haircut” and “Discretionary” at the same time. There’s a tag cloud on the sidebar so you can see the relative ratio of your expenses. Clicking on any of the tags brings you to a good summary page that shows a line graph at the top with all the transactions listed below. The line graph can be put into Yearly, Quarterly, Monthly, (wait for it) Weekly and Daily granularity. So, you can visualize your weekly spending in your discretionary budget.

Almost there, but still some limitations:

  • The budgeting gets the biggest suck – seriously, it’s almost not there. You set a prose goal, say “save up for my hair transplant” and tie tags to the goal. You set whether you want to get to a value or stay under a value. When you get to that level, the goal completes and you start another goal. You can’t do anything recurring and you can’t define any time-based bounds to the goal. There’s also no alert system based on goals.
  • Can’t get the intersection of tags - like Mint, there’s no way to say “show me all the discretionary expenses that are attributed to me”. This makes using tags for sub-categorization difficult. A minor annoyance, but it’s a little less useful. EDIT: Thanks to the comment from Marc Hedlund, CEO of Wesabe, it looks like I missed this in the documentation. You can do some rather powerful boolean searches in your transaction list. You can’t do it yet in graphs, as far as I can tell, but this helps with identifying tags.
  • Tagging could be a lot easier – it takes a lot of clicks. I want to be able to tab through things without my mouse.
  • Finding untagged items is difficult – I don’t mean completely untagged items, though that’s buggy, too. I mean, finding things that didn’t get assigned to one of my five most important categories. The tag filters don’t let you do any logical operations. EDIT: this concern is removed by the more powerful boolean editing that was pointed out in the comments to this posting by Marc Hedlund.
  • Maintaining tags is really difficult – my first iteration ended up with 80 tags. It’s difficult to maintain consistency across all that. Let’s say you come up with a great way to label your haircut. Then, you decide you want to look like an 80s rock star, so you grow your hair long, tease it up, and put your money towards codpieces and hairspray instead. When you get laughed out of the rotary club and get your haircut again, you have to remember that awesome tag from 8 months ago. (You looked like that for 8 months?!?)

Summary: Wesabe is the closest. It has the right visualization, but no good alert system. No good budgeting system. I have to check it myself.

Sumary 2.0: After seeing the advanced search parameters that Marc Hedlund points out, I’m no longer concerned about the organizational aspect of the system. It’s much easier to keep the tag base up. The visualization still needs some work. If Wesabe let me use the powerful searching in its visualization, and let me memorize that view as a report and let me e-mail that report on a user-defined basis, I’d be in heaven. The budgeting stuff still needs work, but I can live without that if I can get regular reports. Wesabe is clearly ahead of Quicken and Mint in this regard.

Conclusion

They’re almost there, but they still suck. I looked at Yodlee, Buxfer, Mvelopes and several others, but they were immediately bad. The three above are the best out there. And that’s depressing. If someone could unite the tagging features of Wesabe with the budgeting features of Quicken and the visualization features of Mint, they would have a super personal finance tracker that would be poised to take over the world. All services lack a really good custom reporting and e-mailing system. None of them have iPhone apps, though they have seriously reduced iPhone optimized websites (that suck even worse).

But I still have hope. I’ve eradicated all my tags from Wesabe and am just using it for my top 5 categories. That will be my weekly budget view. I’m continuing to maintain Mint for the great visualization and spending breakdown, and Quicken for the budgeting. One could argue that I’m doing more work now than I was doing entering all those receipts…and they would be right. This won’t last long, so I’ll continue to hammer on each of the systems to fix their flaws and come up with something truly useful.