Location Based Advertising – an introduction
Recently I was asked to respond to some questions around Location Based Advertising as part of an opinion piece. I thought it would be useful to republish my responses here, and promote some debate on the issue of location based advertising.
What does location based (aware) advertising actually entail? Could you give some real life examples?
Location aware advertising opens up new frontiers for brands to place their advertisements in front of consumers at or near the place where they can obtain them – thanks to the ubiquitous nature of mobile. In the most basic sense, if an ad server is made aware of a user’s location, then a decision can be made to deliver an advertisement based on either the user’s current location, or a history of where they have been, coupled with a user profile of their interests and previous purchases etc.
While the often used “Starbucks example” sounds exciting – you’re walking past a Starbucks and you get an SMS for a free coffee, it is unlikely that this would ever work in practice. Importantly in this example, Starbuck don’t discount their coffee, or have a loyalty program and in cities like London they are simply everywhere.
A more likely use case would be for a movie chain. They most likely will have a movie club and a list of subscribers who have opted in to receive movie news and discounts. If the movie chain wished to fill seats on a slow Thursday afternoon, they would want to know how many of these opted-in members are near one of their cinemas. Importantly, they do not want to know if they are actually at the cinema (then there is no point in sending them an offer as they are already there seeing a movie), but if they are close enough, and the offer is compelling enough (eg 50% off any movie in the next 30 minutes) then they are more likely to respond to the offer – creating a very targeted, and instantly successful marketing campaign.
In the real world today though, the movie campaign example above would be difficult to execute. Let’s assume for a moment that all of the movie club members have GPS. What percentage of them have the GPS turned on (battery life is measured in hours when a mobile phone has the GPS on), and what percentage are not indoors, and hence able to provide a location fix? Also, how often would you need to check the location of each of these users to see if they are close to the cinema? What is the cost to the subscriber, or advertiser for each of these regular location requests, and how many are wasted if most of the movie club members are nowhere near a cinema. The same holds true (and in a more expensive way) if cell-ID is used, as each subscriber would need to be “polled” on a regular basis (at a wholesale cost of an SMS) to see if they are near a cinema.
The most effective and efficient way to provide this sort of LBA is with zone detection, where some software resides on the handset, or the SIM card, loaded with all of the movie chain’s cinemas as profiles. Every minute or so, the software checks locally in the phone to see if it is near one of these zones, and if and only if they are “in the zone” will a message be sent to the ad server. For LBA to take off, it must have an element of zone detection for the push-ad campaigns to work at all.
Starhub in Singapore have just released their own LBA service. Given the way they have designed the service, Starhub may find that in a very short period of time, their cell-ID servers are being overloaded with location requests to check if a user is near a particular store. Checking a subscriber’s location from the network side consumes massive amounts of network resources that are also required to send revenue generating SMS, voice and data traffic. It will be interesting to see how successful this service becomes, and what the return for Starhub actually is.
Location and advertising seem to be a match made in heaven – what’s holding it back?
The ability to provide a user’s location in real time, at a reasonable price and with a reasonable level of accuracy and reliability is something that has eluded the mobile advertising space until now. Issues with the available technologies (cell-ID and GPS) and their suitability for location based advertising not being commercially or technically viable have been the main reasons for the delay.
With cell-ID, the inherent low accuracy of this solution (100’s of metres to tens of kilometres) coupled with the fact that an operator has to charge the equivalent of a wholesale SMS charge for each location request means that the use of cell-ID for LBA will never take off.
GPS on the other hand is being introduced into a number of new phones, but as a technology it is totally unsuitable for LBA because of the issues with high battery drain, inability to provide a fix indoors, and the time for an initial fix mean that for advertising GPS will never be a real contender. What is required is a technology that sits between cell-ID and GPS, and provides a relatively good level of accuracy without the need to add anything to the phone, or consume network resources by regularly checking a user’s location from the network side to see if they might be near an advertiser’s store.
Can location aware advertising exist without the operator?
Yes – companies such as Google have shown clearly that they can provide a reliable operator-independent service. The latest version of Google Mobile Maps (GMM) has a feature called “my location” which allows the user to find their current location on a Google map without dealing directly with an Operator. Google have cleverly deployed the application to enable those users who have a GPS device turned on to report back the cell-identifier (cell-ID) of the nearest mobile base station while the GPS enabled handset is using Google maps.
Google have developed a global database of this information which allows the majority of users who do not have a GPS handset to also pinpoint their location as the GMM application reports back the current, or “serving” cell-ID and this is compared with the Google database and plotted on a map. Reports from many in the industry suggest that Google has developed their solution without any assistance from the operators.
One issue Google are already facing however is the cell-ID database integrity, as operators are known to change cell-ID numbers frequently, so Google’s system will have a lag as new cells are heard and need to be mapped into the database at the correct position. Clearly, the Google my location initiative has proven that location can be provided without the operator, and importantly, without the need for GPS.
The step for Google from placing a blue dot on a map, to pushing location based ads with their significant ad inventory should be a small and obvious one.
What measures can be put in place to avoid intrusion? Are consumers hesitant about location aware advertising?
There are many stories around about the potential for location based advertising to be perceived a SPAM. Hopefully operators, advertisers and marketers won’t just use raw location in their targeting criteria. Increasingly, consumers are becoming aware that the location of their mobile phone (and hence themselves) is now a reality, with the mass market availability of mapping and navigation services.
Early SMS services that sent unsolicited messages have made consumers very wary of mobile advertising solutions. Giving the end user total control to allow their location to be turned off in any application will give them confidence that they own their own location information. Opt-in and opt-out become even more important when location is involved.
So, what would be the ideal conditions for proximity advertising to really take off?
In general terms, there are 4 key areas of focus to ensure mobile advertising as a whole is a success.
1. Flat rate data plans for all. The UK and other markets have made a good start here, but the realisation by other operators that flat or low rate data plans WILL drive mobile advertising is key
2. A Common framework for mobile page rendering and addressing. Since the beginning of 2008, there is strong evidence that the “m.” address is winning over .mobi, but the industry has a have a long way to go on the issue of page rendering to ensure mobile compatible pages are presented to mobile users.
3. Mobile location without GPS, but instead using zone detection. Existing location technologies available to day for mass market location are cell-ID from the network operator and GPS. Network cell-ID systems have not had any significant investment as there have been few positive business cases
4. Customer profiling. Smart collection and use of subscriber information goes hand in hand with location information. Location can be seen as one of many targeting and profiling inputs. If you already know I am a 39 year old male living in postcode W11 in London and I like gadgets and technology, the type of ad I am likely to respond to is fairly well defined. Layering this with real-time location information is icing on the cake as I am most likely to respond positively to an ad that is targeted correctly, and is sent to me when I am close to a location where I can take immediate action on the advertisement.
How do you see the future of location advertising? What will be both the technological and business changes?
For location advertising to really work, more work on the targeting and segmentation of subscriber data, as well as intelligence on user behaviour needs to be undertaken.
Recent initiatives by the GSMA & UK operators to explore delivery of cross-operator metrics for mobile advertising, as announced at the recent Mobile World Congress in Barcelona show the commitment from the operator community to make mobile advertising a success.
In terms of technology changes, new approaches to location around zone detection, where the entry or exit of a pre-defined zone is reported to an ad server rather than the exact location (speeding up the location fix as well as the time to select and serve the ad) will be key technical innovations to power location based advertising.
What standards need to be put in place for the respective players?
In countries such as the UK, there has been a code of conduct for location services for some time, however the existing code may not adequately address new uses for location for advertising. The most basic requirements will be for the end user to implicitly and overtly opt-in to receive location based advertising. The ability to easily opt-out of the same should be key also. Interchange of location information should also be standardised between operator and non-operator groups so that an easy to understand API can be used by 3rd party developers wanting to provide everything from “mashups” to large scale advertising campaigns.