This recent conversation was about the results of the pilot program. The pilot program was a continuing experiment they were conducting with a performance based payment model. Apparently, the program was a success. But it appeared that they were still trying to figure out how to transition the performance based model into something truly profitable for their agency. They were still grappling with a payment model that made sense for both parties and whether to continue on the direction of being a supplement to the paid search efforts of their clients. The opportunity to truly work with the long tail was there, yet how would the metrics work? The larger retailers his agency had serviced were either doing search in house or had another agency or both. Could they truly be profitable with this model and could they grow this model in any real way?
I have my doubts. It's not that I don't believe in the long tail. I believe there are long tail keywords that can be successfully mined. But I don't think it can be the crux of your business. There is just too much synergy between the broader keywords, including brand terms and the longer tail keywords. I'm just not convinced that the two should be separated into two distinct efforts that most likely wouldn't talk to one other. If you are working with an agency, that agency should be controlling your whole program. If your agency is not doing this, than you should shop around for another agency. If you are doing your search in house, there has to be a point where you look at your program and decide whether you can successfully grow the program to the fullest extent it can be grown.
With over 2.5 mil skus at B&N we had a distinct interest in growing our long tail. During my stint there, we made several attempts at casting that wide net to mine our catalog to the fullest extent possible. We were never quite satisfied with the results. We worked with our agency at building out the long tail as well as a search affiliate who worked within strict guidelines about what they could or could not bid on. The attempts were mostly unsuccessful. I'm not saying we didn't find some sweet spots, some unlikely success stories. But as a whole, the effort and expense to find the sweet spot was not worth the return without the coverage of our broader revenue driving terms. We were only able to successfully pay for the experimentation under the umbrella of these other more profitable terms. Without them, I cannot imagine we would have been able to convince our management team that the program was working. I cannot imagine it would have been profitable for our agency either since they worked on a performance based model.
But I do strongly believe in search as a strong customer acquisition tool. I believe you have to look at the bigger picture. If you are receiving a good deal of revenue through your inexpensive brand and broad keywords, you should be able to balance your spend on other less profitable keyword experimentation. To me, it is really an inexpensive alternative to traditional advertising and much more effective at driving brand awareness. Your non brand keywords will be what is driving your customer acquisition and should be treated as a customer acquisition spend. You have to decide how much a new customer is worth to your business. It is always a balancing act and search is a moving target.
I would encourage this agency to continue the dialogue with prospective clients. If their customer acquisition efforts are worth these tests they offer perspective clients, then they should continue this effort. But I would encourage them to make the case for the holistic management of their client's paid search efforts. If they are truly experts at the long tail and they are a good service agency, then there is no reason why they won't be able to make headway in this area.
