Do you really have to create AutoCat rules for every transaction?

New to Tiller, love spreadsheets. HOWEVER, it seems crazy to me that there are not a set of standard categorization rules native to Tiller. Several examples come to mind:
-For all grocery stores, shouldn’t there be a general rule that categorizes the transaction as “grocery”?
-For the same dollar amount that debits your bank and credits your credit card, shouldn’t Tiller be able to match these and categorize them as “Transfer”?
-For gas stations, shouldn’t that be auto categorized to gas?

I see the real value of Tiller once categorization is setup. However, seems too much work to setup bespoke categorization rules.

While some transactions may be relatively simple some are more complex than first glance. Ie for the categories. You can add a column called “Category Hint,” that I believe Heather suggested elsewhere that you can have the AutoCat system pull off of? but say you have costco charges, is that for gas, for groceries, or other items.
For transfer that assumes you have both banks set up, and tiller can’t always assume that’s the case.
I believe there’s a long thread of how other categorize their own setup to help you see how others have categorized their items.
There’s also a couple posts where users have shared a couple of scripts to help build the rules of what you’ve previously categorized. auto cat also has the " Rules from Past 90 Days" you can test as well

One of the things I love about Tiller is precisely that it doesn’t assume anything about how I want to categorize something, unlike other budgeting software. It’s very easy to set up an Autocat rule.

Yeah, I get where you’re coming from, but that works better when categories are pre-set/fixed.

With Tiller, categories are fully customizable. I might categorize a grocery store transaction as “groceries” or “food”. The same goes for “Transfer” or “gas”. Someone might use “Fuel”, instead of “gas”, and use “gas”, instead of “natural gas”.

Tiller’s Philosophy

Tiller does not automatically categorize transactions by default, but we do offer an automation tool, AutoCat, for both Google Sheets and Excel to help speed up your categorizing workflows.

We believe manually categorizing your spending builds a clear picture of where your money is going. Besides, most other tools we’ve tried that automatically assign categories fall short. How can they know the intent behind your $30 purchase at the coffee shop? Perhaps part of the purchase was a gift, so why would we assume?

To sum it up here’s why we take this different approach to categories:

  1. It’s quick. You can categorize your spending in just a few minutes a few times a week. It takes up less than 1% of your time.
  2. It builds awareness. You know exactly what you spent and where when you touch each transaction. You’re more engaged with your money this way.
  3. It aligns your spending with your intent. As previously mentioned, we can’t assume to know your intent behind any purchase.
  4. It’s more accurate. When you’re doing it yourself, you always get it right.
1 Like

thanks for all of the replys. I made it through all of the categorizations, and now set up to use on a regular basis. However, I think there are some things that would be great to have for new users:
-examples of “standard” autocat entries. For example, there are standard gas stations that are known, and could be copied and pasted by users into their autocat tab. Same goes with groceries and restaurants, although they are more regional.
-A way to highlight unreviewed transactions. As I have been using it for the past few weeks, I find that new transactions show up, they get auto categorized, but I would still ke to know they are in the right buckets. Could you highlight “new” transactions as they get added to the transactions tab?
-transfers between accounts should be automatically paired. if I pay a credit card from my bank, it should show up as a transfer on both sides, and net zero on the budget.

Great app, this is my financial tool now!

It’s obviously Tiller’s prerogative to have their own philosophy, but I just think this is silly.

I love being able to make my own categories and have flexibility over how things are categorized, but if 10/10 times I categorize a transaction that looks like X to be “Food”, there is absolutely no need to make me set up a rule to do that, or manually do it every time. Computers are pretty smart. (See Tiller AI AutoCat).

Let me review the 1% that are new/confusing, or correct special cases (this shopping purchase was really a gift or whatever) but don’t make me do the work to do it 99% of the time.

In my humble opinion the main thing getting in the way of Tiller being used by 10X more people is the out of the box experience to just get started and get something useful is pretty painful. Lots of set up for rules, minimal default reporting that just makes your data come to life, etc.

I’m a big believer in the potential here (otherwise I wouldn’t have invested 100s of hours over the last month in building stuff for free for this community) but I think this philosophy is counterproductive for usage & growth.

I agree the initial work was painful. I wish it would have canned AutoCat options (ones that you can see and edit is fine!), a way to identify new and questionable transactions, and a way to automatically categorize transfers. It would make onboarding much easier.

1 Like

Interesting discussion. I’m torn:

On the one hand, just today, I’m getting ready to take the plunge on @cps’s cool AI Autocat tool. Because I think it’s very cool, and I want to see how well it works.

On the other hand, a couple of weeks ago, after about two months of using Tiller, I summarily deleted all of my Autocat rules. Every single one of them. I felt like I was missing things in my transactions and my spending habits by not categorizing things manually.

I guess the conclusion is that there isn’t necessarily one right answer to this, but I suppose it’s better to have the option and choose not to use it than not have the option at all.

1 Like

I get that, I guess I’d like both ways:
-system highlights new transactions for you since last time
-autocategorizes “basic” ones (but using a rule that is editable)

to manually solve the first one, i started highlighting transactions that I have reviewed. Then new ones should be easier to see.

Right. But even there, it gets tricky. If I see a $50 charge at the Sheetz, then I know it’s gas (Auto & Gas). If I see a $7 charge at the Sheetz, then I know it’s my teenager buying candy (Snacks).

Hey, @cps , can your AI tool tell the difference between those two? Repeated pattern of same merchant with different size transactions that go to two different categories.

Right now I’m not passing amount in, but it would be interesting to try that as an extension and see how well it works.

A few notes…

  1. As far as “examples of standard autocat entries”, we do have the Rule Generator in the TMF add-on for Sheets. That is a great way to get started and get to know the process. Perhaps this is not well advertised.

  1. Many users use filters to view uncategorized and recently-categorized (hint: add a “Categorized Date” column). Would this work for you?

  2. I agree that more automation around transfers is as important as automation on categorization— especially since it affects cashflow calculations. There are some challenges here though like the timing of the two transactions and also when one side of the transfer is an account not linked to the spreadsheet.

This is a great discussion everyone. Keep pushing us— especially you, @cps, with your ML tool!