FAQs on Price Change Analysis

Reading Time:

The Price Change Analysis insight in Sprk™ provides a deeper understanding of your customers’ price sensitivity. This leads to better informed price decisions in the future that maximize prices while minimizing customer push back.

How is this analysis done?

For each price increase, the sales of the product are analyzed for 182 days (91 days before and after the price change). Biases in the data are removed to ensure the analysis is as accurate as possible. If the results are statistically valid, they will be shown on the Insights page 92 days after the price increase.

For businesses new to Sprk™, the analyses for price changes that occurred between 3 and 4 months ago will appear right away.

Why doesn’t this analysis appear for some price increases?

To ensure statistically significant results, at least 30 units of the product must be sold in both the before and after periods. Low-volume products don’t have enough data to draw any meaningful conclusions. In fact, it’s very easy to draw false conclusions from sparse data.

In addition, products are excluded if the price was changed more than once in the 182-day period. Multiple price changes make the analysis extremely complicated. They also don’t give customers enough time to adjust to the new prices, which is critical in avoiding the perception that you’re always raising prices.

How is seasonality removed from the analysis?

Almost every business has a seasonal effect on sales. If you increase the price at the end of your busy season, you will naturally have more sales before the price increase than after. This can lead to a deeply flawed analysis because it’s unclear whether sales are down because of seasonality or the price increase.

To remove this bias, we adjust the product sales based on the overall seasonality of your business. This involves comparing the total sales from your business for the “before” period to the “after” period. This is the seasonal adjustment for your business that we then apply to the product-level metrics.

For example, let’s say your overall sales decreased 10% between the before-and-after periods. If the sales decreased by 15% for the product with the price increase, 10% of that is attributed to seasonality and 5% to the price increase. This is an imperfect technique, but it’s better than ignoring the seasonal effect. The only case where it is invalid is for products that have a different seasonal pattern than your overall business.

Do large orders impact this analysis?

No, unusually large orders are excluded from the analysis. For example, let’s say you’re a bakery and you raise the price of cupcakes. If you receive an unusually large order (such as for a wedding) in the “before” or “after” period, that order is ignored to remove biasing the analysis.

Are there any cases where this analysis is not valid?

As stated above, the analysis is not shown if it’s not statistically valid. However, the app cannot distinguish between products that are not available for sale from those that did not sell. If you have products that have sporadic availability or are only available during a specific time of year, the analysis may be invalid.

For example, let’s say you offer smoothies only from May through September. If you raise prices on June 1, the “before” period would have sales for only one month, and the “after” period would have sales for three months. In this case, the analysis should be ignored.

Please contact support@getasprk.com with any additional questions.