Online Versus Offline Prices:
Is There a Difference?

Many shoppers assume that online prices are lower than in-store prices. However, a recent study by MIT Sloan Prof. Alberto Cavallo found that prices are actually identical about 72% of the time for products sold both online and offline. According to Cavallo, consumers shopping at multi-channel retailers, which sell both online and in physical stores, will find the same price whether they shop online or at any of that retailers’ physical locations.

While this is useful knowledge for consumers, it’s also very important for economists who are increasingly using online prices in measurement and research applications. They do this even though less than 10% of all retail transactions in the U.S. are conducted online.

To provide clarity on this issue, Cavallo conducted the first large-scale comparison of online and offline prices in 56 large multi-channel retailers in 10 different countries. This type of retailer still concentrates the majority of all retail transactions. He excluded online-only retailers like Amazon and Ebay from the study, as they currently represent a small percentage of retail transactions in most countries.

The online and offline data collection was carried out as part of the MIT Billion Prices Project. For the offline prices, Cavallo used crowdsourcing platforms to hire over 370 workers tasked with scanning barcodes and collecting prices for a random set of 10-50 products in any physical store of a given retailer. All workers used a special BPP app developed specifically to simplify and standardize the data collection process. Once collected, prices were sent to the BPP servers and compared with the online prices for those same items.

“The results were surprising to many people. There is a general belief that online prices are different, but no agreement on exactly how they differ. This study showed that prices are quite similar in these large multi-channel retailers” says Cavallo.

He found that online and offline prices were identical about 72% of the time with more similarity in apparel and electronics and less similarity in other sectors like drug stores and office supplies. “This is likely because customers can see the price of a TV at each store location and online so it would be hard to explain any price differences. As a result, someone who is living in a remote location, where transportation costs may be higher, pay about the same price as someone living in a large city.”

However, the percentage of similar prices does vary somewhat by country. Cavallo found that the similarity rate ranged from 42% in Brazil to 91% in Canada and the UK. The U.S. was the closest to the average at 69%. “It’s possible that retailers in countries like Brazil think online buyers have higher incomes and can afford the higher prices – or the delivery costs is built into the price,” observes Cavallo, noting that the degree of price difference appears to depend on the state of development of a country’s offline and online markets.

He adds that he saw no evidence that retailers try to obfuscate price comparison by changing the products’ ID numbers or price discriminate using “dynamic pricing” strategies. In particular, it was clear that retailers in the U.S. are not varying their prices with the location of the IP-address or when the scraping robot repeatedly browses the same webpage of a particular good for longer periods of time.

“These results are significant for economists and statistical agencies, as they show that online prices are similar to offline prices. The ability to cheaply and quickly collect – and use – massive amounts of online price data provides unprecedented opportunities for economic research. The findings also imply that online prices can be used for consumer price indexes,” says Cavallo.

He notes that the takeaway for consumers is that they don’t need to make decisions to buy online or offline based on price. Instead, they can rely on other factors like convenience, speed of purchase and product selection.

Cavallo adds, “For those interested in the effects of the Internet on retail prices, my study suggests that while the web may not have reduced price dispersion across different retailers, it may have created incentives for firms to price identically across their own stores.”

Alberto Cavallo is a professor at MIT’s Sloan School of Management.
You can find a more in-depth account of his research here.