New data show that conversion rates are down among social media users across some top social platforms (Facebook, Pinterest, and Twitter). The decreased rates seem primarily the result of increased visits (more visits and the same number of sales will obviously lead to lower conversion rates). But what’s wrong with this study?
The data used to show the decrease in conversions comes only from Branding Brand, a single mobile commerce site provider. Although Branding Brand has a lot of high-profile retail clients (CostCo, Bath & Bodyworks, Crate & Barrel, Timberland to name a few), it’s still only one mobile commerce platform and as such its data do not represent results across the entire mobile ecosystem. Additionally, the data used represent only traffic from Branding Brand sites, not all web traffic or even all mobile traffic for a particular retailer. Comparing Branding Brand data in context with complete traffic data for a retailer would tell a more interesting story.
The fact that Branding Brand is the single platform for all of the data presented in the study may show more about the type of traffic that works for platform itself than social media or mobile in general. As one cynic commented, “The data is just saying that BB builds mobile sites that don’t convert.” While mobile conversion rates were actually up overall (and down only for social media), we don’t know the full situation until we compare the Branding Brand data to data from another mobile commerce provider (or, preferably, several). Additionally, we don’t know much about the type or quality of the social media traffic considered in this study. Are users clicking on specific deals from the Facebook brand page and not converting (perhaps because a deal has expired), or just making general exploratory visits to the retailer website? Knowing more detail about the type of social media sourcing would allow for more actionable conclusions.
Fan and Follower Bias
This data jives with some of my previous assertions about the value of a fan. Sure, fans can be valuable, but the willingness to become a fan suggests a preexisting brand affinity that biases the fan group. While limited due to its sourcing in one platform, the Branding Brand data about the low conversion rates for Facebook confirms my more general impression: that people are on Facebook to socialize, not to find brands.
Better Luck Next Time
A common problem with social media “studies” is that they tend to be performed on data from just a few brands or agencies, which creates an automatic bias and complicates the extrapolation of the results to other businesses or industries. Just as, say, medical studies in a specific population can’t necessarily be extended to people with different demographic characteristics, studies of specific brands or platforms can’t necessarily be extended to all organizations. The Branding Brand data is not particularly flawed, but it is incomplete, not necessarily applicable to all brands or platforms, and does not offer enough detail to drive specific business decisions about where to allocate resources.