In a recent report called The State of Marketing Data prepared by Netprospex over 60 million B2B records were analyzed in order to assess the quality of the data measured as fitness for use related to marketing purposes.
An interesting find was that out of a score of maximum 5.0 duplication, the dark side of uniqueness, was given the average score 4.2 while completeness was given the average score 2.7.
This corresponds well with my experience. We have in the data quality realm worked very hard with deduplication tools using data matching approaches over the years and results are showing up. We are certainly not there yet, but it seems that completeness, and in my experience also accuracy, are data quality dimensions currently suffering more.
In my eyes the remedy for improvement in completeness and accuracy goes hand in hand with even better uniqueness. It is about getting the basic data right the first time as described in the post instant Single Customer View and being able to keep up completeness and accuracy as told in the post External Events, MDM and Data Stewardship.
Liliendahl on Data Quality 