Reputation: 1
I'm currently looking at daily sales data of a specific brand over the course of the past year. My objective is to create a formula to roughly estimate the sales growth for future months.
My project isn't going very well, as the brand is very volatile in monthly sales, making it impossible to predict with a basic linear formula. I'm arriving at the conclusion that a single year's worth of sales isn't enough data, and I may have to result to provide a specific formula depending on the month. Is there anything I haven't thought of?
Note: Recording of sales start on the 15th of every month
Upvotes: 0
Views: 59
Reputation: 563
Your sales are showing seasonality. Consider using sasonal ARIMA models.
Upvotes: 0