MSearchGroove

In-brief: The excitement at last week’s Mobile World Congress may be about location-based communications services and advertising, but Andrew Grill, General Manager of Sales & Business Development at location technology provider Seeker Wireless, argues that basing the delivery of contextual content & advertising on our traditional understanding of location is not enough. He offers another solution that covers all the bases.

When Peggy asked me to write a guest column for MSearchGroove, I jumped at the chance, as it gives me an opportunity to correct many of the myths surrounding location technology and make a case for zone detection - a concept that will take targeted mobile advertising to a new level.

The blogosphere is buzzing with news and views about how easily location can be combined with other demographic and customer specific information to provide a more targeted proposition for advertisers. However, the fact that location services have not yet crossed the chasm is proof that providing a location element to mobile advertising is a lot tougher than we think.

As I suggest in the title of this post, location technology has yet to really make its mark on mobile services. In short, location has been “lost” for some time, discredited by hype and hyperbole since the late 1990s. As an industry, we have only ourselves to blame.

For a start, we have unquestionably accepted the false assumption that location is really just about GPS (Global Positioning Systems) technology. This rather simplistic view is supported by the spread of personal navigation devices and popular Sat-Nav (Satellite Navigation) systems selling like hotcakes in mainstream shops. Another boost comes from Nokia, which has introduced a slew of GPS-enabled devices such as the Nokia N95, Nokia 6110, and, more recently at Mobile World Congress, their “pedestrian navigator”device - the Nokia 6210.

But let’s not forget that GPS has been around for some 30 years. It’s hardly new news - and it’s not suited to deliver the range of location-enabled services the industry promises are on the immediate agenda. Put simply, GPS is excellent for outdoor navigation, but severely challenged when used indoors. What’s more, the battery life of a mobile phone with the GPS turned on can be measured in single digit hours. The most I get out of my Nokia 6110 is just three hours with the GPS turned on. Clearly, these practical considerations and limitations will slow the uptake of location-enabled services such as mobile advertising, for which it ironically holds the most promise.

And we have to ask ourselves just how valuable GPS information is in the first place. Clearly, location as a concept is a perfect fit with GPS-enabled devices and delivery schemes. (I am here and I can see a spot on the map.) However, being able to provide information on a user’s coordinates - and a visual representation to match - is hardly a key enabler for the delivery of relevant or useful mobile advertising.

If my GPS device tells me that I am at 51.507778N, 0.128056E, what does that really say about me? Granted, the information tells an advertiser that I am in London. But London is a big and busy city. Does it therefore follow that I am in a location in London, say near a cinema, that would make it worthwhile for the advertiser to show me a relevant ad? Possibly. But to know this for certain the advertiser must process this location data further, and it’s a process that takes time - time that could be spent serving me an ad!

To recap: GPS is accurate - if used outdoors, if the user has it switched on, and if the handset has the capability. Since we cannot assume that these three key criteria are met, we cannot rely on mobile advertising based on GPS to deliver the best results.

Network Cell-ID, a location technology available in most networks today, doesn’t represent a silver-bullet solution to this dilemma. We would expect that Cell-ID uses triangulation to work out where the user is, utilising multiple cells and time differences between signals from the towers. However, the reality is quite different. Mass market network Cell-ID systems use just the current or “serving” cell - hardly an approach that guarantees the level of accuracy advertisers expect.

And it gets worse. Because mass market location-based services in Europe have been stuck in the bowling alley, operators have not invested heavily in scaling capacity in their Cell-ID servers. There is a serious supply bottleneck when it comes to location data. In fact, I’m hearing that location throughput can be measured in single digit queries per second! Think that one through. To complicate matters, I hear mobile operators are still wrestling with the issue of how to grant secure, trusted access to their subscriber’s location.

The delivery of effective mobile advertising doesn’t require the system to know the exact location of a subscriber; but it does demand that the system knows where the user is relative to some point of interest (shopping mall, cinema, restaurant). This information helps advertisers deliver ads in tune with user location context and arrive at the all-important decision on what content to provide (or not provide) to a user based on their location.

This is where zone detection comes in. Zone detection places the decision to trigger an advertisement at the handset rather than the network. Seeker Wireless uses patent pending algorithms to instantly determine if a subscriber is in a particular zone and then send a message to a server if there is a match - all without the need for GPS. It effectively and instantly qualifies a user as either in or out of a zone, providing a binary result that can easily interface into an ad campaign manager.

How does this work? Let’s consider a requirement to detect if a user is indeed standing at the ticket office of the London Eye, without explicitly asking the user where they are (or if they are in London at all). Zone detection asks the handset to report what it can “see” in terms of the immediate radio environment and, based on this input, decides if it does indeed recognise the “landmarks” corresponding to that zone.

Zone detection thinks this through, connects the dots, and arrives at a binary decision as to whether the user is inside or outside the specified zone. A targeted message or advertisement can be sent if and only if the user is in the relevant zone - much in the same way that Google AdWords only charges the advertiser when the ad is actually clicked.

Imagine the profound impact zone detection could have on mobile search and advertising. White label search engines such as Medio and JumpTap are pioneering the use of relevant customer profiling to ensure the right content is delivered to the right user at the right time. Now the way is clear for them to add location to the mix to achieve even better results. Indeed, adding the user’s relative location via a binary zone detection mechanism is likely to be a very powerful proposition moving forward.

It could also turn out to be quite profitable for all parties involved. Consider this email I received from an executive, which outlines the downside (and the prohibitive costs) associated with a network approach to location information.

“…We are currently stuck in the nightmare of waiting on carriers to decide how and when to offer location data and when they are they are charging too much per transaction…”

This plea for help says a lot (and I do see many emails like this).

It is indicative of a problematic issue surrounding all network-based approaches that does not bode well for an industry eager to harness location to enable a new breed of mobile advertising services.

Indeed, a network-based - or carrier-led - approach gets even trickier if we consider that global brands (the ones with the money to ultimately fund mobile advertising start-ups) want reach, and will likely reject approaches that reach only a subset of the mobile users in a country. The U.K., for example, is home to five mobile operators - most with their own location servers and feeds.

There are companies, such as Mobile Commerce, that aggregate LBS feeds, but they must charge a commercial rate - measured in pence/cents per location lookup - to recover the cost of using the mobile network to initiate a location request in the first place. To locate a phone, similar resources are used as for sending an SMS - so this must also be charged. In short, a network-based approach to location is not only fragmented and complicated - it’s also quite costly. A network-based approach is also flawed because an advertiser that wants to push location-relevant advertising must constantly poll (or locate) the handset to know where the user is at that moment, on the offhand chance that they are indeed in the right place at the right time.

Let’s take the Starbucks coffee example, which everyone in the industry (myself included) has used at one time or another to illustrate the power of location-based advertising. You’re walking past a Starbucks and… you know the rest.

However, Starbucks is probably the least likely to use location advertising. Why? Because Starbucks is everywhere - sometimes on the opposite sides of the same street. There are 178 within 10km of central London - and 171 in Manhattan alone (see 171starbucks.com). Starbucks charge a premium for their coffee and don’t discount or have a loyalty card - so why would they pay 5-10 pence per customer to regularly poll handsets just in case a user might walk by?

The bottom line is there is no mobile advertising business case that could support a per subscriber location charge on a daily basis just to see if a subscriber is near a store.

The rumour I hear in the industry is that one European mobile operator is gearing up to launch a location-based mobile advertising solution based entirely around GPS. (And broadcasting company CBS in the U.S. is set to follow suit.) My response: Think it through and consider the end-user environment when choosing the appropriate location technology. As I have pointed out in this column, GPS is not a de facto fit with location-relevant mobile advertising campaigns.

In contrast, zone detection brings relevance to location and enables profitable location-enhanced mobile applications. Granted, the technology has been around for some time, but its advance has been slowed by issues such as how to accurately and reliably provide detection capability on the handset. With that issue now solved the way is clear for some exciting applications.

Leading the pack is zone detection for mobile advertising. Expect to see some examples of real and significant progress in 2008, and perhaps even some live campaigns. Indeed, the pieces are falling into place to make zone detection more than a prime-time technology; it’s poised to dethrone GPS as the solution essential to ensure the effective delivery of context-relevant mobile advertising campaigns.

For more thoughts on mobile advertising, mobile location and local search visit Andrew’s blog at blog.andrewgrill.com

Andrew’s next guest column will explore location-based advertising, with a special focus on Google’s operator-independent modus operandi.

• Posted in
• Location-based services, Mobile Search, Mobile Advertising, Local Search

13 comments

share

These icons link to social bookmarking sites where readers can share and discover new web pages.

• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 

permalink