I Asked an AI App Every Question About My Spending. Here Are the Results.

One of the promises of AI expense trackers is that you can just ask questions about your money and get answers. No spreadsheets, no manual filtering, no exporting to a pivot table.

After a month of logging in DrakeAI, I put this to the test. I asked every question I could think of about my spending. Here is what actually happened.

The setup

I logged consistently for 30 days - about 8-12 transactions per day, a mix of food, transport, subscriptions, and occasional larger purchases. At the end of the month, I had roughly 280 expense records covering most of my spending.

Then I started asking questions.

Questions it answered well

"How much did I spend last month?"
Answered instantly with a total and category breakdown. Accurate to what I had logged.

"What are my top three spending categories?"
Good. Showed Food, Transport, and Subscriptions with amounts. Matched my intuition.

"How does this month compare to last month?"
Showed a side-by-side comparison. My transport spending was down because I worked from home more. Food was up. This was genuinely useful.

"How much have I spent on subscriptions this year?"
Pulled all subscription records and totaled them. Fast and accurate.

"What did I spend at restaurants last week?"
Specific time range, specific category. Worked correctly.

"What is my average daily spend?"
Calculated from the logged period. Gave the number with context (highest day, lowest day).

"How much did I spend on coffee this month?"
This impressed me. It pulled all Food records mentioning coffee, plus the Starbucks transaction I had not tagged as coffee specifically. Total was right.

Questions that worked partially

"Am I spending more on food than I should?"
The app gave me my food spending total and percentage of overall budget. But "should" depends on income and goals I had not set. The answer was useful data, not a judgment.

"What is my biggest waste?"
This required the app to make a value judgment. It surfaced categories where I was spending more than average and flagged subscription renewals I had not used. Useful starting point, not a complete answer.

"Show me what I spent on work expenses."
I had been tagging some things as "business" in the notes but not consistently. The query pulled what it could find. The gap was in my logging, not the query.

Questions it could not answer

"Am I on track for my savings goal?"
The app does not know my income or savings goal. It tracks expenses, not income. This is a budget-planning question that requires more setup.

"Should I cancel Netflix?"
A decision question, not a data question. The app told me how much I spent on Netflix and how often I had noted streaming expenses. The decision was mine.

"What will I spend next month?"
Projection questions require historical pattern recognition that goes beyond what the free tier offers.

What I learned

The Q&A feature is strongest for retrieval - pulling specific numbers and comparisons from your history. It is weaker on evaluation and judgment, which is expected.

The insight I did not expect: asking questions changed how I logged. Knowing I could ask "how much on coffee?" later, I started logging coffee purchases more carefully rather than just including them in "food."

The Q&A loop created a feedback effect on the logging quality. Better questions led to more useful data.

Who this is most useful for

People who want to understand their spending without learning an analytics interface. If you have ever thought "I wonder how much I spent on X last month" and then not bothered to find out because it would take 15 minutes of spreadsheet work - this is for you.

The answers are seconds away. The data is yours. You just have to ask.

Try DrakeAI free on Android - iOS coming soon.