Last month, an article in the Economist caught our attention. This article covered recent research surrounding eBay’s paid search campaigns, and suggests that the impacts of paid search can be “overstated.” On the other end of the spectrum, you have Google stating the opposite is is true: That paid ads provide clicks you would not get from organic search. Who is right? Let’s look into it further.
When looking at the full eBay research (found here), you learn that eBay came to three conclusions about their paid search campaigns:
1. Sponsoring branded keywords is ineffective and cannibalizes users who would eventually visit eBay via other channels.
2. Non-branded keyword sponsorship is ineffective on average.
3.The small positive effect is for new and infrequent users.
We believe these are great takeaways for eBay, but we would be cautious of taking eBay’s conclusions and automatically applying them to other businesses.
eBay rightly notes in its research that the research is directed towards well-known brands (e.g., Amazon, BestBuy, etc.). For a small or new brand, these conclusions are unlikely to be applicable. However, eBay does believe that its research applies to these other well-known brands and questions why they spend so much money on “a rather ineffective marketing channel.” We find this statement to be problematic for a few reasons:
1. Unusual Business Model: eBay is in a distinct consumer experience, primarily selling unique or used items via a peer-to-peer auction model. This is very different from other well-known retail brands that compete directly against one another selling identical products (Walmart, Target).
2. Competition: eBay notes the data in its research did not include competition for branded keywords. In our experience, it is common for competitors to run ads for branded keywords, and this can be one of the more effective uses of AdWords sponsorship.
3. Keyword Strategy: eBay does not go into detail in its study, but it could be argued that its keyword strategy for paid search is flawed. We see their methods for non-branded search as a “carpet bombing” approach, targeting an extremely broad array of keywords, many of which have no connection to eBay whatsoever. More discussion on this topic can be found here.
The conclusion we draw from the eBay study is not that AdWords is ineffective for all businesses, but that it was ineffective for eBay, and it could be ineffective for other businesses as well. We have long felt that there are examples of AdWords cannibalizing organic search traffic, particularly for branded keyword searches in which the business has a #1 organic placement. On the other hand, perhaps a better keyword strategy could have improved eBay’s AdWords performance.
We hope that businesses will take this research not as a final word but as a trigger for their own experimentation. Our suggestion: Try the same tests eBay did. Find out if your own AdWords campaigns are providing value. If they aren’t, then you can cut them just as you would any other ineffective marketing channel. Good luck!