In a previous entry, I argued that companies should not evalute a product manager's performance flatly in terms of product revenue. How should, then, a company evalute a product manager's performance?
One way that Steve Johnson mentions on the productmarketing blog is in terms of the number of on-site visits to prospective and existing customers. The Pragmatic Marketing folks advocate basing bonuses on the number of on-site visits. I believe that on-site visits (in particular, one-on-one interviews) with customers are the single most effective way for a product manager to gain an understanding of the market. I have observed also that companies tend to underestimate their value.
However, I don't believe on-site visits quite get at performance. On-site visits are a means to an end. That end is an in-depth understanding of the market that a product manager successfully communicates to developers, sales, and marcom. If developers, sales, and marcom then do their jobs right, the product will likely succeed.
A company should evaluate a product manager's performance in terms of how well she communicates an understanding of the market to the development, sales, and marcom teams. I propose that a company can measure this performance by surveying the staff of those departments, asking them how well the product manager has imparted information about the market. Ask sales how well the product manager has profiled the different buyers. Ask development how effectively and convincingly the product manager has communicated the requirements. Ask marcom how clearly and credibly the product manager has defined the key messages to use in marketing programs.