Musings on mobile advertising and targeting of ads

This post was stimulated by Alan Patrick’s Mobile Advertising…….Immobile industry?  In turn, he was responding to a piece by Damian Green on the Telco 2.0 blog.

The first half of his post lists a number of reasons why “mobile advertising will prove to be a sideshow to the main event”.  When I read the list of reasons (and I will summarise in a sec.) I see them as reasons why the sector will be slow to develop rather than reasons why mobile advertising will remain small forever.  From an entrepreneur/VC perspective this makes the question one of market timing rather than market size, and the rapid growth of Admob and Screentonic tell me the market is happening now.  That said these are valid points and will impact the development of the market.

Damians reasons for thinking that mobile advertising will remain a sideshow are (and this is my possibly over concise summary):

  • Mobile is complicated – multiple devices, networks techonologies, network policies – this will make it expensive to produce mobile ads thus shrinking the market – this is true for now, but will become progressively less so over time
  • Advertisers don’t understand mobile and the channel may not prove effective – again fair comment today, but there is enough experimentation going on at the moment to convince me that someone will work it out

The second half of the post goes on to say that mobile operators could leverage their knowledge of consumers where consumers surf and their position of trust to serve highly targeted ads.  This turns the internet model on it’s head, as the service provider chooses the ads, instead of the site owner.

This is a powerful idea, but not necessarily a new one.  I remember having conversations on this theme with respect to the wired web seven or eight years ago, yet it hasn’t happened (yet).  The difficulties on mobile will be the same as they are on the web:

  • Practical difficulties with the service – a user will need to be able to edit his profile – so he can avoid a situation whereby, for example, the operator serves an ad for a holiday in the Carribean when his wife is looking at the phone, thereby ruining his surprise holiday.  Or a gambling advert revealing he hasn’t really stopped betting on the footie, or whatever it might be.  To make matters more complicated the profile should consist of clickstream data plus other factors like age, income bracket, hobbies etc. which the user needs to be encouraged to enter and maintain.
  • Value chain complications – for this to work site owners will need to switch their ad serving partner from e.g. an Admob to the mobile operator – and unless all the operators start doing this at the same time a site owner will suddenly be trying to use e.g. Screentonic for three quarters of its traffic and e.g. Vodafone for the other quarter.  Further, given that mobile operators are also portal operators the notion of trusting the advertising on a mobile site to a company that operates a rival site might prove difficult for many.
  • Chicken and egg – the difficulty with starting any advertising business is that it is hard to get inventory until you have advertisers and vice versa.  This problem will be particularly acute in this model as the inventory will be restricted to a fraction of the traffic on any given site (although I suppose traffic on the operator’s own portal will all be available).

And that is without mentioning the difficulty of working with operators!!!