Artilium

Friday, June 26, 2009

Measuring Relevance in Mobile Advertising

The latest mobile advertising figures show revenue of £6 billion by 2014 (i). This has taken the current economic conditions into consideration and so these are quite conservative compared with some previous figures. However, this is still predicted to be a massive market and is experiencing 38% growth in comparison with a decline in other advertising sectors. The big question is no longer, will we accept mobile ads. Many people already regularly receive mobile adverts and the mobile advertising industry is definitely real and growing. The new questions are more about how they should be delivered and how can they be made more relevant. There are a number of ways they can be delivered to mobile including:

• SMS and MMS
• Applications
• Mobile TV
• Idle Screen
• Mobile Internet
• On-Portal
• Personal Messaging

We have previously discussed how to make Ads relevant through advance behavioural processing and so we now consider the following question. How do we prove they are relevant and how does Artilium address the question of driving and measuring relevance?

In the mobile advertising world there are a number of metrics used to measure things like: number of ads served, response rates, costs and effectiveness. Essentially these are the same metrics used for Internet advertising. The number of impressions is measured in cost per mille (CPM) which is the cost to display the advert 1000 times. This is fine for brand awareness but is not very useful in determining the relevance of a particular advertisement to an individual. An alternative is to measure the response rate, via the number of clicks on the advert link relative to the number of impressions of the advert that were served.  If an advert is relevant you would expect it to be clicked more often than one which is totally irrelevant. The advertiser must somehow pay for the adverts that are served, but is only really interested in ones where there is a click through. These are the ones he is willing to pay for which leads to the pay-per-click model. The industry has adopted cost-per-click (CPC) pricing where an advertiser might compete for a customer based on what they are willing to pay for relevance keywords. E.g. a motor insurance advertiser might be willing to pay 75 cents for a click when a subscriber is looking at mobile content about “insuring your car”, but would only pay 50 cents for a click when the subscriber is looking at “used car prices”.

Artilium’s behavioural processing is designed to understand the customer and their current context so this is much more than just keywords. It stands to reason that the better the customer is understood, the more relevant an advert served to that customer might be. In terms of CPC we have a mechanism to increase CPC by making adverts much more relevant in real-time – right ad to the right customer in the right place at the right time. The increased relevance model is a win-win-win scenario. The customer only receives adverts which are of interest and may actually receive less as a result. The advertising platform provider gets a higher income through an increased CPC and by serving less irrelevant ads. The advertiser might pay more for each click, but the ads are only served to highly relevant customers so the conversion rate will be higher and ROI will be increased.

The win-win-win scenario can only be achieved through more relevance and that is at the heart of Artilium’s real-time Behavioural Processing developments.

(i) Mobile Advertising Delivery Channels, Business Models & Forecasts 2009-2014 (Second Edition), Juniper Research, June 2009.