If recommendations are so important, why don’t companies invest in making them more effective?
In a recent survey by Adobe, 40% of ecommerce respondents have automated recom-mendations on their sites, or are in the throes of implementing them. Roughly 1 in 5 media and financial services companies are investing in automated recommendations.
Yet, half of companies don’t do anything to optimize the recommendations. And a quarter “optimize” by manually updating recommendations. Behavioral, profile, and business rules are each used by just under 20% of respondents to improve recommendations. The takeaway for me is, 75% of companies aren’t committed to making sure a visitor sees the best selection of available content.
Why would a company invest in a technology that has demonstrable bottom line impact, but not tailor it to produce the best possible results?
a. Mediocre (not optimized) is good enough
b. Testing is too hard
c. We don’t believe testing is a good investment, or we don’t understand it
d. We don’t have the resources (yet) – either skills, budget, or technology
I think the answer is a combination of a and b. Once automated recommendations are deployed, most companies get at least mediocre results from them. Recommendations service providers do a pretty good job of automatically optimizing recommendations, and guiding their clients to deploy good (best?) practices. Testing to optimize recommendations requires investing in an ongoing program, requiring specialized and scarce skills. The path of least resistance = set and forget.
On Target 