A paper from the National Bureau of Economic Research concluded that the ROI of search engine marketing are “a fraction of conventional estimates” and in the studied company “become very negative.” How can we reconcile this with what our clients believe and the fact that they’re spending millions with Google?
Their approach was to partner with eBay and selectively turn off certain markets’ Adwords spend and monitor click volume. It’s a reasonable approach and the science is sound, but the last leap from the scientific results to the business implications is where it falls apart.
The first line of the discussion section tells it all: “The results of our study show that for a well-known brand like eBay, the efficacy of SEM is limited at best.” We’d agree that this may be true for eBay, but that doesn’t hold up for all “well-known brands.” Why? Quick- name an online marketplace for used goods. Now name second one. Is that harder or easier than naming two places to buy cell phone coverage?
I’d be willing to bet that the situation changes significantly for more competitive markets. The other factor is that branded search terms are so much cheaper than non-branded that the calculus changes- even if it were only 10x as effective as non-branded it would still make sense.