The duration of transmitted extracts is around 7 seconds for chart hits and about 15 seconds for new releases. The extracts are transmitted in MP3 format in both online and offline versions.

Each evaluation of a track is followed by a measurement of how well the track is known, i.e, its familiarity. Medianalyst can therefore measure the appreciation and the familiarity of 4 tracks per minute, or 360 tracks during a survey lasting one and a half hours.

The values recorded require no transcription or encoding, thus avoiding source of error.

For each evaluation, and for each target considered, Medianalyst provides two main pieces of information, which can be exploited individually or combined :
• the average appreciation
• the range of votes

In radio, the range of values recorded can be an indicator as to whether a track has an overall positive impact or spans a broad scale of appreciation; given two musical items with the same average mark, the item with the less extreme values will be the lowest risk option for the radio station. A music program can therefore be considered as a combination of tracks selected according to their appreciation and their potential to satisfy a wide or limited audience.

A third essential indicator is the familiarity of the tracks broadcast by a radio station, as this factor influences how an item is appreciated and the durability of its success. The combination of appreciation and familiarity enables a track's burnout rate to be measured. Medianalyst can directly calculate adjusted appreciation scores according to a track's familiarity.

These 3 values can be applied to all programming strategies.

Packages of musical items can be selected by year of production, author or nationality, for example. Each package can be classed under a specific denomination.

The marks recorded have no predetermined value and can be expressed in various forms: score out of 10, out of 20, index with a base 100, percentage, etc.

All the data from Medianalyst music research can be imported into traditional music programming software.

Large, widespread population samples, recruited by applying quotas, can be questioned online within a very short time.

Within an identical framework, it is possible to combine or select data from several offline tests, and to consider the results from several sessions.