NetDynamix responds, moots industry standard for internet radio numbers

The company at the centre of an internet radio readership row says it is engaging with industry experts to set up a forum to verify internet radio statistics, and that it issued a letter of demand to blogger Shaun Dewberry because it was aggrieved that he had not engaged with them on his findings.

In an official statement, the company says it does not agree with the methodology or the basis of Dewberry’s arguments and they stand by their stats.

The audio streaming company further indicated it would be interested in getting involved with the development of a “standard” that could be used to define audiences on internet radio.

On Monday, Dewberry wrote a blog post calling the listener figures of South African online radio stations, including 2OceansVibe Radio and Ballz Radio, “complete fabrications. Utter nonsense. Lies, even”.

NetDynamix claims it has been investigating the possibility of bringing on an external auditor to “confirm the accuracy” of its stats. Since Dewberry’s allegations emerged, the company says it has been engaging with “independent industry experts” who will check the accuracy of its numbers.

According to NetDynamix boss Hanz Stricker the company “assumed the ‘sessions’ were listeners” and was “not aware of any standard”. He added that the company has been trying to get current unique user numbers, but that its service provider “has not been forthcoming”. Stricker did say, however, that the company was applying a retrospective analysis on the unique user numbers.

The audio streaming company explained that it had served Dewberry with a letter of demand because he did not approach it for comment before publishing his blog post. Had he done so, it claims, it would able to show him that his research was incomplete because he did not have full access to all the company’s data or an understanding of the methods used to determine statistics.

In his blog post, Dewberry examines the figures published from NetDynamix via Shoutcast, a piece of software that relays and distributes streams to listeners. The majority of radio stations that use the software are listed on the Shoutcast site and Dewberry says these figures are much lower than those publicly claimed by the internet radio stations.

Every public Shoutcast server publishes their listenership to this directory. The largest have 12 000 listeners maximum at any time. If the biggest stations in the world only have 12 000 listeners, how can two start-ups (2OceansvibeRadio and Ballz) in bandwidth-starved Africa have tens of thousands more listeners than that?

Memeburn notes however that Dewberry appears to take issue with the internet stations’ concurrent user numbers (users listening or who are on a website at any one time), where as NetDynamix claims to provide its clients with numbers based on “sessions”. The company defines a session “as a connection to the server(s) in order to receive the audio/video stream”.

“As with any research related to defining audiences there are many variables to consider with collating data and the interpretation of this data. NetDynamix, on request, provide their clients with numbers based on “sessions”. A session is defined as a connection to the server(s) in order to receive the audio/video stream. This does not take into account that a listener could change devices or lose their connection to the server, at which point it needs to be reestablished, or the change in IP address of that connection or change in route from point of connection to breakout onto the internet; to name just a few examples of multiple sessions being initiated to the service,” said NetDynamix in a statement.

This method of collecting listener numbers does not take into account that a listener could change devices or lose their connection to the server, at which point it needs to be reestablished, or the change in IP address of that connection or change in route from point of connection to breakout onto the internet, which NetDynamix says could result in multiple sessions being initiated to the service.

The company says that it would not be in its best interests to give its clients “any information other than the raw numbers generated by these ‘sessions'”.

“Our retrospective analysis shows a 10% discrepancy in between ‘sessions’ and unique numbers,” Stricker said. He seems to think this isn’t a massive difference, but when you’re dealing with the massive numbers NetDynamix was claiming, it patently is”.

2OceansVibe today said that it “in no uncertain terms” thought that an audit of NetDynamix was necessary and that it stands behind the creation of an industry standard.

Quite what this standard would entail is unclear. World Wide Worx CEO Arthur Goldstuck, “the nature of online radio makes it more difficult to measure audience numbers”.

“It’s not just about the number of streams, two or three people in one office could be listening to one stream”.

Goldstuck reckons that it may well “be necessary to apply traditional radio measurement standards”.

Terrestrial radio in South Africa is measured by the South African Advertising Research Foundation (SAARF) through the RAMS methodology.

Under RAMS a representative sample group is given radio self-completion diaries to complete in their homes over a seven-day period. That’s a very low tech solution, but at least it’s an established standard.

FULL PRESS RELEASE BELOW:

PRESS RELEASE REGARDING RECENT ALLEGATIONS AGAINST NETDYNAMIX

In response to the recent statements made against ourselves (NetDynamix) and our clients (Ballz Radio, 2oceansVibe Radio and Kingfisher FM) we would like to clear up a few points.

NetDynamix is a technology supplier to the radio broadcast industry and our core business is the distribution of our clients content over the internet. At present we service 73 clients on a regular basis, some of these being FM/AM stations that simply relay the terrestrial broadcasts on to the internet and the balance are internet only radio services.

The technology and means used to distribute our clients radio stations is not something that we are willing to make public as this enables us to deliver a cost-effective solution to our clients where others find it difficult. We believe that this is not an unreasonable statement, as the technology in question does not contravene any laws or codes of conduct. We therefore suggest that the figures quoted in the published statements are not a reflection of the bigger picture as the general public do not have access to all of our systems. Prior to the issue being raised on the internet, NetDynamix started investigating the possibility of finding an external “auditor” to confirm the accuracy of our numbers. This is still part of our strategy and we see the importance of this “certification” process as an added benefit to our clients, and the industry as a whole, and we are more than happy to engage with the correct parties going forward. In the meantime, NetDynamix has identified a few independent industry experts, who we will be engaging with today, and will setup a forum to verify our stats. The process has already started and the names will be released as soon as the people identified confirm their availability.

We do not agree with the basis and method that Mr Shaun Dewberry has used to arrive at his findings simply because he does not have full and unhindered access to our entire network, nor does he have an understanding of the methods that we employ to manage our network and it’s resources. We are further aggrieved that 1) we were not invited to engage with Mr Shaun Dewberry during his process of researching his article at which time we would have been able to illustrate to him that his view was not holistic, and 2) the defamatory nature of his article towards NetDynamix following the release of his article.

As with any research related to defining audiences there are many variables to consider with collating data and the interpretation of this data. NetDynamix, on request, provide their clients with numbers based on “sessions”. A session is defined as a connection to the server(s) in order to receive the audio/video stream. This does not take into account that a listener could change devices or lose their connection to the server, at which point it needs to be reestablished, or the change in IP address of that connection or change in route from point of connection to breakout onto the internet; to name just a few examples of multiple sessions being initiated to the service.

It would not be in the interest of NetDynamix to give their clients any information other than the raw numbers generated by these “sessions”.

The following is a random day in May 2012 (which is not their highest figure set available) for Ballz Radio which shows the hour by hour total number of sessions which were initiated to our servers for the Ballz Visual Radio Stream:

08:00 – 6 012
09:00 – 8 408
10:00 – 15 359
11:00 – 26 579
12:00 – 48 599
13:00 – 49 337
14:00 – 59 764
15:00 – 57 364
16:00 – 54 359
17:00 – 46 666
18:00 – 10 953
19:00 – 8 121
20:00 – 8 137
21:00 – 2 338
22:00 – 2 102
23:00 – 2 666

On average our systems reflect that the audience for Ballz Visual Radio is composed of 31% International listeners and 69% Local.

NetDynamix have no reason to inflate these numbers as the clients are not remunerating us on a per listener basis, they pay a fixed monthly fee in order to use our system. We standby what we have presented and reiterate that we are investigating a way to independently verify these figures and methodology via an impartial technical specialist, auditor or journalist.

Similarly 2oceansVibe figures have also been raised. Their session statistics show a different session spread, typicaly known as a trend, but essentially make up a similar number of sessions at particular broadcast hours.

On the matter of Kingfisher FM, we cannot entertain a comparitive view of a partial month of figures (the month that they moved to us) coupled with a total peak in that we do not have access to their listenership figures prior to them moving their audio streaming service to our company, as this was not provided to them by their previous hosts.

NetDynamix would be interested in getting involved with the development of a “standard” that could be used to define the audiences on internet radio. We encourage robust debate on this and other subjects related to internet radio.