Directions Magazine (DM): For many years, operators seemed to ignore LBS as a revenue opportunity. What’s changed and what do you deem the catalyst for this change?
Michael Flanagan (MF): In our view, it’s not so much that operators ignored LBS. It is more that the opportunity was seized more quickly by so-called “over the top” (OTT) providers. These players were able to use techniques like GPS and cell triangulation to derive a user’s location and provide very specific information to those individuals. And once those links were established, the monetization opportunity came much more quickly for them.
The “big data” phenomena that has gripped the industry more recently means operators are now realizing that there are new opportunities to monetize location information based on the data assets they have immediately at their disposal. Importantly, these are assets that are unique to operators, and not something OTT players can muscle in to without formidable effort.
DM: Can you better describe “Location Insight Services ”? Is this a new term specific to operators?
MF: It’s not a new term as such. The “Location Insight Services” (LIS) market already exists but has so far had very little attention because until now, it has been much harder to do.
Location insight services use aggregated and anonymized “trend” data from connected consumers’ mobile location data. They are distinct from location-based services (LBS), which are dependent on the availability of individual real-time data.
Location-based services usually allow service providers to know where an individual is at a specific point in time, and for example, offer those individuals locally relevant services or offers, or use location as validation (e.g. payment).
This differs from location insight services, which allow service providers to know more about where groups of individuals are over a certain time period. This may also include whether those groups are travelling through a certain location or are static. This allows service providers to have a greater degree of insight into people’s behavior, and allows them to predict future behavior patterns.
LIS are incredibly valuable because they use trend data collated over time to build insights about locations, and support decisions relating to those places and individuals within them.
DM: The press release mentioned that STL Partners identified two business models open to operators to realize the value of LIS. Can you articulate the business model for both the standalone and partner option? Would the operators or partners sell directly to the OTT and what kind of location-based data would they be selling?
MF: Operators have two principal business model options to realize the value of LIS.
Firstly, by enabling API access to aggregated datasets, operators can provide “raw,” unanalyzed big data for integration by partnering organizations into other data platforms. These partners can then conduct their own data analysis to their own ends. An example here might be a big retail chain using location data analysis to establish new retail sites.
Alternatively, operators can partner with an in-house department or external organization that conducts data analysis and interpretation on behalf of other businesses. An example here might be an advertising agency that provides consumer insight to its own clients.
An operator probably wouldn’t provide data to what we’d consider a traditional OTT as those players tend to be focused on providing specific services to individuals; LIS data are best applied to situations where partners want to capture significant location trend information about groups of individuals – by age, by gender, by period of time, etc.
DM: Also mentioned in your report are the challenges of big data and an eight-step guide. Can you share some of the advice you would have for dealing with the volume and velocity of location-based big data?
MF: One of the biggest hindrances to establishing revenue-generating location services has been a lack of clear business cases for them. The intention for this report was to take away some of that mystery and shed some light on what those business cases might be and how much they’re worth. Beyond this concept, the eight-step guide is primarily a suggestion about how operators can release the value in location insight services:
1. Start with your own business and technical operations and consider the project as an integral part of your own big data challenge.
2. Prepare a fully-costed business case.
3. Select a proven technology solution.
4. Adopt an agile development approach.
5. Clearly articulate a differentiated value proposition for your clients.
6. Collaborate on industry standard definition of “location.”
7. Establish a business critical metric (“interesting” insight is not enough) for clients.
8. Don’t force your (new) clients to change their practices to use your service.
DM: Once the business model is established and the product identified, the report suggests that operators will offer aggregated, anonymous location data with a time component. What level of geographic accuracy are you suggesting operators can provide given that many apps today are offering location data based on geofencing or other narrowly constrained geography?
MF: Most LIS solutions today offer cell-level accuracy (that is, the user location is estimated to be at a spot within the footprint of the cell providing service). JDSU LIS solutions offer improved accuracy with better than cell-level resolution. In many important applications, the accuracy is building-level. And, in contrast to many apps with distributed/individualized location functionality, LIS is centralized and aggregated in a manner that traditional LBS apps would struggle to realize.
But location is actually only one component in the provision of LIS. There are several underlying technology requirements that support the extraction of meaningful data that can be analyzed. Bearing in mind the requirement in LIS for high precision trend data – not just location data – the following requirements exist:
• High user (subscriber) coverage (e.g. devices, apps and user activity)
• High geographic coverage
• Low device impact (e.g. battery life)
• Good data continuity, 24x7
• Low-cost “big data” acquisition or harvesting
• Speed /granularity of data capture
• Location accuracy and granularity (location resolution below cell level is significantly more valuable)
DM: One of the opportunities mentioned in the report suggests you can provide competitive benchmarking. How are you attributing your data with visitor profiles or other data typically based on loyalty programs specific to the retailer? This would suggest that you are adding some demographic or psychographic data. How do you intend to provide this?
MF: Benchmarking is a good application of LIS in the retail sector. The idea is that retailers, or retail planners, can establish location insights about store visits of target customers in their own stores versus those of competitors.
For example, a retailer addressing the youth audience with youth fashion could use LIS to identify which competitors’ stores in a shopping mall are visited by that demographic group, and how often, on what days, in what order and so on. Using that benchmark, the retailer can measure the impact or effectiveness of a new marketing or ad campaign.
Importantly, the “loyalty” aspect of the benchmark is derived from the location data from the mobile devices, not from loyalty programs of the individual competitive retailers being benchmarked.
DM: The press release mentions that the “power is back in the hands of the operators” to monetize “mass location intelligence.” Isn’t this counterintuitive to today’s need for precise positioning? Perhaps you can better explain “mass location intelligence” and how it will be monetized.
MF: The requirement for precise positioning and mass location intelligence are not mutually exclusive. Mass location intelligence is about understanding, in a comprehensive and holistic way, the movements, needs, trends and behaviors of users at particular locations (or sequences of locations). To do that, you need to capture precision location data over periods of time, and be able to aggregate those data and analyze them effectively. So precision location data, as previously discussed, is but one of the elements required to build and interpret “mass” trend data.