top of page

Using Modern Data Science Tools to Predict Revenue from QuickBooks


Last month, we released a tool that essentially serves as a playground for using Meta's "Prophet" modeling tool. You can find it live here: https://prophet-revenue-projection.streamlit.app/


Here's a quick guide on how to use it with your QuickBooks data. Please note that we do not retain any of the information that you submit, which can be checked at the open source repo here: https://github.com/Depot-Analytics/Prophet-Revenue-Projection


Generate a transactions report


1. Login to QBO and go to the Reports Tab



2. Enter "transactions" into the search bar and click on "Transaction List by Date"


3. Change the date range to your desired time range to use as baseline data


4. Filter by the type of transactions you're interested in predicting


5. Export the report


6. Open and Reformat the CSV


  • Delete the top 3 rows with metadata

  • Delete the bottom row with metadata

  • Rename the "Amount" column as "y"

  • Rename the "Date" column as "ds"

  • Format the "y" column as numbers


7. Upload the CSV to the revenue projector

  • Click the "Browse files" button and upload your reformatted CSV



  • Change the parameters to reflect the realities in your organization!



Are you interested in further using modern data science tools to project do forecasting in your business? We'd love to help - setup an appointment or send us some information for a free consultation at info@depotanalytics.co!

bottom of page