READ BACK
ISSUES ONLINE

Mobile Advertising: targeting the unknown

Jo Wall, Proposition Manager, Mobile Advertising, Acision
 
The promise of mobile advertising is that it enables brand to consumer advertising which is based on a direct relationship with an individual instead of merely a contextual or broadcast one.  There are 4.1bn mobile phones globally compared to only 1.4bn TVs and 1.6bn internet users (Mobile and Internet figures: Internet World Stats.  TV figures: Nationmaster.com).
 
The ability to directly target billions of pre-qualified consumers with tailored messages through a more measurable and interactive channel is an advertiser’s dream!  But is it turning into a pipe dream? Despite the promise of mobile advertising its full capability is not a reality today.  Yes, it’s true that mobile is the most measureable of mediums.
 
Each call or interaction produces separate pieces of user information that can be analysed to provide behavioural targeting segments. Yet with complaints about invasion of privacy and mobile operators not visibly willing to undertake data mining activities to extract the digital oil of the 21st century will the personal mobile channel become nothing more than a small screen to the unknowns?
 
There are a number of factors which contribute to the operator’s reluctance.  Namely, the operator’s back office servers are not built as a cohesive system making the correlation of data harder; the operator’s revenue from mobile advertising is very small compared to other services such as messaging so is it really worth their while but, more importantly, there is always the fear of a subscriber backlash – especially in markets where the subscriber penetration is reaching or has exceeded 100 per cent.  Undertaking activities which could result in churn is not an acceptable risk in the operator’s eyes.  Does this lack of targeting data mean that mobile advertising is deemed a failure before it has truly launched?
 
Not necessarily.  There are many different ways to obtain this precious targeting information and not all of them require the mobile operator’s network data.  However, for mobile advertising to flourish it is critical that privacy, permissions and preference are at the heart of any solution going forward.      
 
What is preventing targeted mobile advertising taking off?  One area for concern is that of privacy.  On 13th January 2009 the Centre for Digital Democracy (CDD) and the US Public Interest Research Group (USPIRG) filed a complaint with the US Federal Trade Commission (FTC).  They urged an immediate investigation take place into mobile advertising to stop the rise of “growing invasion of consumer privacy”.  A US survey conducted by BIGresearch of over 22,000 consumers in June 2009 seems to support the basis of the complaint.  The survey revealed that the number of people who think that mobile marketing is an invasion of privacy has increased to 52.1 per cent from 49.5 per cent in 2008.
 
There are three aspects to this.  Firstly how the data is obtained, secondly the level of trust which surrounds the organisations extracting the information and lastly the control and use of that extracted data.  The one common denominator from an end user’s perspective in all of this is the unknown.  This unknown caused a furore in 2008 when BT secretly used customer data to test Phorm’s advertising targeting technology.  A report based on a recent survey carried out by University of Pennsylvania and Berkeley Centre for Law & Technology found that a large majority of US internet users reject the use of behavioural advertising – 66 per cent of adult US citizens do not want advertising to be tailored to what advertisers think are their interests.  Once it was explained how this information was derived e.g. actual methods used to track behaviour, this figure rose to 73 – 86 per cent.
 
This concern flows onto the organisations performing the analysis.  There are a numerous companies claiming to be able to extract the targeting information but what security regulations surrounds them? Would it be more transparent if there were fewer organisations responsible for this activity but the security and policies which they adhered to were more transparent, regulated and monitored – alleviating the unknown?  When it comes to use of the data, a key aspect is having the reassurance that an end user can easily opt-out and more importantly that all your associated data is purged.  This element of control will encourage continued participation in mobile advertising and future campaigns.
 
But of the “unfair and deceptive” practices, such as behavioural targeting in the CDD/USPIRG complaint – isn’t that a step too far?  After all other industries such as the retailers seem to have more freedom, e.g. supermarkets regularly analyse buying behaviour for audience segmentation.  Tesco, the UK’s largest supermarket chain, has a loyalty scheme.  Their Club Card holders give Tesco permission to analyse their buying habits and other information in return for points that can be redeemed against their shopping bill, or for air miles and other deals.  This has enabled Tesco to come up with over 100,000 customer segments to improve its selling efficiency.  You could argue the difference here is that Club Card holders have given their permission, whereas some of the analytics being undertaken on mobile maybe unknown to the end user.
 
The mobile phone has become such an indispensible part of our lives this, combined with the measurability of the channel, may mean that so much more can be learnt about your whole lifestyle rather than merely your grocery list.  Not everyone would want person2person text messages analysed to allow for relevant ad insertion based on key word searches?  Hence the need for permission, including agreeing to what information can be analysed.
Which leads us onto preferences: Contextual advertising based on the type of programme, what time it is aired or type of magazine is the basis of advertising for print or TV advertising.  Why is such contextual advertising not deemed sufficient for mobile?  Is it a question of the targeting needs to reflect the personal nature of the channel?  For mobile internet banners contextual advertising is acceptable but for push messages, because of the very fine line between relevant campaigns and spam, targeting is deemed essential?  Analysis of operator network data can reveal your location, what websites you browse and length of text can indicate gender – females apparently send longer SMSs.  A level of preference can be surmised from these facts.   What remains unknown is, for example, whether the end user is browsing automotive sites because they are looking to buy or sell a car.  It is only by engaging with the end users will marketers be able to have this deeper level of insight.
 
The US Federal Communications Commission (FCC) has effectively put a stop to any use of network orientated data for marketing purposes.  The FCC view a text message to be the equivalent of a call – any analytics on a text would be akin to wire-tapping.  As part of the entrepreneurial spirit targeting information or preferences are instead obtained via SMS/MMS dialogues with the end users.  But it could be equally achieved via a section on an operator’s or brands portal allowing subscribers to add their demographic and preference information.
 
However, some analytic companies are of the belief, similar to Dr Gregory House from the US medical drama – House, that “all people lie”.  They may say that they are interested in sports but in fact their browsing behaviour, for example, doesn’t support this.  What should be used or even which is more valuable to the advertiser?  Today, the most sought after marketing attributes are generally location, age, gender, type of handset and if possible salary bracket.  In most cases, when advertisers are asked as to how valuable  additional targeted information is worth, for instance what CPM increase could be obtained knowing whether someone is an F1 fanatic , most will not provide an answer.  That is because it is currently unknown.
 
As an industry we have to be vigilant that we’re not deliberately setting the bar for mobile advertising so much higher than for all other mediums. The mobile phone is the most personal of devices and quite rightly caution is needed.  However, analytics does not have to equate to sinister or suspicious activities.  Without some level of insight the mobile could be transformed into a broadcast channel to the unknown.  In return for this insight we must honour privacy by gaining permission to determine preferences.  It must be easy to consent (opt-in)and withdraw permission (opt-out) and there must be transparency about the scope of activities.  This enables the end user to share the information they feel comfortable with.  What does this mean for advertisers and mobile operators?   For advertisers it requires a more collaborative approach, openly engaging with their consumers to solicit their interests.  For mobile operators it depends on the level of involvement they are willing to undertake.  They do hold a rich vein of subscriber information which if extracted or solicited sensitively can increase their role in the mobile advertising value chain by making such targeting data available.  However, if and how they do this, within regulatory boundaries, is up to them.  What of those operators who do nothing?  One could predict that they will be relegated to a mere dumb pipe.  But the truth is unknown.