ADOTAS – One question that’s often asked is whether energy prices impact e-commerce or other online transactions. While it’s a well-known fact that people drive less as gas prices go up, the impact of higher energy costs on online behavior isn’t well understood.
Does the squeeze of higher gas prices reduce online purchases, as consumers cut back on discretionary expenditures? Or alternatively, do cash-strapped consumers buy more online as they shift spend away from activities that require driving?
In a recent study, we looked at the impact of energy prices on consumer behavior. Our findings indicate that higher energy prices likely change online behavior and could be good news for Google. For more information on this subject, as well as, specific insights about your industry, download our Q1 2011 Paid Search Research Brief.
Energy Prices Correlated to the Number of Paid-Search Clicks
In looking at data over the last five quarters, we found a strong correlation (R-Squared = 0.97) between the price of oil and the number of paid-clicks. If this data is broadly representative, it suggests that changes in oil prices and the number of paid-clicks are strongly correlated and move in similar ways. Our data sample was drawn from the Marin Global Search Index, which consists of 800+ clients managing over $2 Billion in annualized ad spend.
So what does this mean… are paid-search clicks really tied to gas prices at the pump? Our analysis suggests that the answer might be yes. That being said, it’s still difficult to draw predictive conclusions because we only looked at five quarters and our data set was a small, albeit meaningful, subset of Google’s total click-volume. At the very least, businesses should include changes in online consumer behavior when they model high energy-cost scenarios.
Whether or not high energy prices are good for Google gets a bit more complicated. If our data is broadly applicable, then Google’s top line could benefit from higher energy prices. But a big part of Google’s fixed expenditures come from the power costs of running large data centers. And as power prices rise in tandem with energy costs, Google’s bottom line will likely be impacted by a more expensive kilowatt hour. Net net, it’s difficult to say how this would affect Google’s bottom line.
Methodology and Statistical Background
To measure the relationship between energy prices and online behavior, we calculated the coefficient of determination between oil prices (i.e. the avg price of a barrel of crude oil) and paid-search metrics like impressions, clicks and costs.
The coefficient of determination, also known as R-Squared, explains how well a statistical model fits the actual data. An R-Squared value of 1 implies a perfect fit between the statistical model and the underlying data. An R-Squared value of zero implies the opposite. In other words, the closer R-Squared is to 1, the stronger the relationship between two data sets.
This article was originally published on Marin Software’s Marketing Insights blog on May 2. Reprinted with permission.