Excerpted from Steve Yelvington’s [sober, sage and real] media weblog
Counts are not the same thing as surveys.
Surveys yield projections that have margins of error that should be disclosed and explained.
Survey methods should be disclosed and critically examined. If a survey was conducted in order to generate a press release for marketing purposes, it’s probably bullshit…
Pageviews are not the same as hits.
Unique users are not the same thing as unique visits.
Attempts to count unique users are universally inaccurate because of multiple device usage and cookie clearing.