The stampede to new Google products and metrics for evaluating product success

Paul Kedrosky’s post about how little Google spends on advertising and promotion reminded of something I heard Google’s Marissa Mayer say in her talk at the WebGuild 2006 conference this morning. Marissa said that when they evaluate the success of a new product, the first thing they do is ignore the first month’s worth of data. Because of the mad stampede (those are my words, not Marissa’s) to new Google products, the first month’s worth of data is almost always outlier data.

The other part of what she said was also interesting — that they look for month over month growth of 20-30% as an indicator of strong success and that when m-o-m growth is in the 5-10% range, they take that as an indication that they aren’t doing enough with the new product, that there’s not enough consumer affinity to the product.