BPR considers that online research (web-enabled) offers a potential solution to a number of obstacles facing traditional methods of data capture. Key benefits of web-enabled research include the ability for respondents to self-complete tests in their own time, convenient data retrieval, an alternative method of obtaining participation amongst “hard to get” demographics and, in some cases, cost savings. BPR however strongly cautions radio stations against thinking that online testing produces more accurate or reliable data for the purposes of strategically determining a radio stations format compared to traditional forms of research such as CATI or face to face. Online research is not immune from the reality of requiring a market representative sample and it is in this area that the accuracy of online research faces its most important challenge, as do all forms of research capture. Good research stops and starts with interviewng a statisically representative sample of the target.

The principle application for online research in the radio industry at present is for music testing.

The testing of music online is a highly viable alternative to call out and particularly auditorium music tests (AMT‘s), BPR considers the traditional AMT to be particularly irrelavent with a highly questionable accuracy. Online music testing also provides the opportunity for cost effective “fun tests” within a radio stations database community creating the perception of listener involvement in the music played by the radio station.

Online music testing is yet to demonstrate that it can deliver better music research than CATI, particularly when it comes to determing music playlists for AC stations. In fact, comparative testing confirms that in order to achieve the same representative accuracy in online testing compared to call out, stringent guidelines must be followed when conducting online music testing.

With regard to determining a radio stations playlist music by using an online system BPR is specifically concerned about two dangerous paradigms; these being: A. It‘s OK to have only a P1 view of the music and B. Using a station branded website to recuit and/or retrieve the music test data. BPR believes that a P1 only view of the music risks shrinking an AC stations audience rather than expanding it, while a station branded website creates a bias and generally attracts the more extreme music fans which often manifests itself as songs on the fringe of market taste starting to test. This will often be regarded as great news by station Music Directors because it creates the impression that online music testing has “discovered” new songs the station can now play, when in actual fact all it is doing is telling the station to play music that the greater majority of radio listeners either don‘t know or don‘t like.

For a niche music station, a P1 only view of the market may be fine, particularly if the station‘s music niche is unique.

While the relative guidlines for online testing may differ slightly from country to country depending on internet usage and social issues, BPR recommends the following checklist for any competive radio station wishing to grow audience beyond its current cume.

Know Your Stations Target

Selection of respondents for a music panel should only be done when the precise target of the station’s music format (strategy) has been determined. This will generally be defined by a P1/P2, age, sex and music preference criteria.

Ensure Representative Recruitment

Radio stations should recruit respondents to a music panel in a way which is randomly representative of the market. In most markets recruitment via the telephone remains the most effective way of achieving this. A combination of telephone and online is also OK provided all respondents are filtered by the same panel criteria.

Panel Maintenance

BPR recommends a program of panel maintenance based on regular email and personal telephone contact. Panel size should be sufficient to allow a 3 week panel rotation if required. It has also been found that a reward/incentive may be required for on-line testing to keep respondents active.

Test a Mix of Music

It is important to always test a mix of music genre and style to ensure “panel skew” is avoided. For example only testing currents will cause 80’s preference listeners to slowly drop out of the panel. Stations should test the breadth of music within their music strategy. This is particularly critical for AC stations.

Ensure a Robust Sample

An effective sample of 100 to 120 respondents should be the goal for each test. This will probably require around 180 invitations.

Use The Advanced Analysis Tools Within BPR Xplorer

Xplorer offers a number of great tools to analyse your music data. Format Correlation and Song Compatibility are both excellent for optimising playlist decisions.