How to ask an assistant to find historical patterns in my monthly utility bills

"Historical pattern" implies more than one year of data — a genuine seasonal pattern (higher gas bills every winter, higher electric bills every summer from AC use) only shows up when you compare the same calendar month across multiple years. A single year of bills can look "spiky" without revealing whether that's seasonality or a one-time event, so the first step is making sure you actually have enough history before asking for a pattern analysis.

Finding the real seasonal signal

1. Gather at least 2 full years of billing history. One year can't distinguish "this always happens in winter" from "this happened once" — two or more years of the same calendar month gives you a real baseline.

2. Ask the tool to group by calendar month, not just list bills chronologically. A prompt like "group these bills by month name (January, February, etc.) and show the average cost per month across all years" produces the actual seasonal comparison.

3. Request a bar or line chart, not just a table. A chart of average cost by calendar month makes a seasonal curve immediately visible — a table of the same numbers requires more mental math to interpret.

# Grouping by calendar month across multiple years reveals seasonality
import pandas as pd

df = pd.read_csv("utility_bills.csv", parse_dates=["bill_date"])
df["month_name"] = df["bill_date"].dt.month_name()

seasonal_avg = df.groupby("month_name")["amount_due"].mean().sort_values(ascending=False)
print("Average bill by calendar month, highest to lowest:")
print(seasonal_avg)

If you only have a few months of bills, be honest with yourself (and the assistant) that you don't yet have enough data for a seasonality claim — a real pattern needs repetition across years, not just a couple of adjacent data points. The right response with limited data is to keep tracking and revisit this analysis once you have a full year or two on file.