If you are in the business of making killer apps, then MobileDevHacker was the place to be in Boston this week - where an all-star lineup congregated at the NERD Center in Cambridge to hear, mingle, and discuss topics that drive serious growth. This was a collection of topics that confront app executives and developers alike, to see how the industry attacks some of the toughest challenges of the past, and the future bets they'll be making.
Some of the key speakers/attendees included Havas Media, App Annie, TUNE, AdColony, Rovio, Fiksu, The TapLab, M&C Saatchi, and many more.
Skyhook’s VP of Marketing, Mike Schneider, spoke on a panel with Aubriana Lopez, Lead Strategist for Wireless, Hyperlocal and Mobile at Digital Element. The panel covered "How to Use Deep Analytics Around Geofencing", and was moderated by Microsoft. Here were some of the main points covered:
How to get started
1) What pieces of information and infrastructure are required to perform analytics around geofencing successfully?
You will need the following data: time, precise lat/longs, and contextual data surrounding the location - or venues visited. You will also need the infrastructure - meaning geofences that can scale.
Today, there are limits on how many you can do - only 20 geofences at a time on iOS and 100 on android. Skyhook’s Infinite Geofencing, you can scale an infinite amount of geofences, regardless of device. This will help you to rapidly deploy and report on a limitless number of geofences instantly with point-and-radius or polygonal boundaries.
If you want to understand where your customers are, the venue piece is an important part of this and we have a unique client-side way to make them infinite. You also need flexible campaign controls - adjustable geofences based on customer-defined zones (geofiltering), which gives you the ability to upload custom venues.
2) Using the collected of information, what type of insights are you looking and how do you apply them?
You need to do some aggregation by location types or by brands to make it useful. Any app with location enabled can extract this data in the form of location requests. Then overlay the lat / long log information with a venue database to understand in aggregate where your users go.
This can be a laborious, manual process, but on-device software can automatically deliver the contextual data you need. Location-based context will allow you to categorize these places so that you can aggregate information by things that make more sense than looking at usage by each individual coffee shop, retail stores, airports and on and on.
For some examples, take a look at your apps’ usage:
- Where are people spending the majority of their time at what time of day?
- What type of people are spending time at which venues?
- At what locations are your users engaging with which of your apps’ functionality?
Once you have this location data, you can then roll up places by Brand (Dunkin Donuts, Starbucks, CVS, Walmart, etc.) or Category (retail, pharmacy, coffee shop etc.) instead of trying to make sense of each address or location point.
Compare your Vital Ratio (which is defined as your total app usage / visits) across these locations and you can see if users are getting the value you expect or if they are finding the functionality they need in the place that they are. Understanding this will help you prioritize how to design for place.
4) Having data to analyze is good because you can begin to make predictions on future trends but being able to deliver those predictions in a timely fashion is critical. What is the approach you take to deliver the right content at the right time?
Consumers make it clear what they like, where they are and what they are looking for in online conversations on social media. Now, with mobile ingrained in all aspects of a consumer's life, mobile app usage can tell apps a great deal about users including what they like, where they go, and when they use the app. The mobile phenomena has opened up a plethora of opportunities for you to gather data on your consumers. How to leverage this data is the next question.
You need a balance between the list of features your app displays with when and where they are useful to your users. Not every feature in your app is useful to your users at all times, in all locations. There’s an interesting trend emerging now where apps are slowly starting to incorporate “in-store modes” into their functionality. Retail apps such as those offered by Walmart and CardStar use intelligence gathered from their users based on location to best serve them the functionality they may need. For example, when you enter a Walmart store, your entire app experience shifts to a “Store Mode” that lets you take advantage of coupons, deals, and other relevant store information to help you make decisions while shopping.
5) How do you handle situations when there is not a location signal?
Depends on what data you have. If you have IP location, you can generate a reliable lat/long based on the partner you choose to work with - like Skyhook.
6) Looking to the future, Are there any technologies you see influencing or changing the way you using geofencing to and deep analytics to deliver an even more personal experience?
While it’s great to know what features your users are engaging with, there’s more to the story, which is why we stress the importance of looking at where your users go by getting a high level view and then diving down. You can do this by segmenting your users into high, medium, and low usage categories to find out where the power users are and to compare that to others. Then look at the functions they use by place to get a sense of what’s important in the places they go. Looking at ratios of usage over total number of visits (your Vital Ratio) tells you how vital that function is in that place.
For example, if we see that our Auto Intender persona accesses the payment functionality of our app 70% of the time when in a retail location. If we rank our payment functionality as one of the most vital features of our app while in a retail location, what can we do to the design of our app to get that number up to 100%? The answer may be to enable a “shop mode” that is triggered on the device by a geofence when a user enters the store, and have that functionality displayed prominently so users don’t need to dig through screens to pay for their purchases.
This level of contextual data is critical to delivering the most relevant and valuable experiences to users in real-time, as well as driving global optimization of an app to suit the entire audience better. Dynamic user experiences can respond to users’ intent, adapt to their location, change based on time of day, and tune to their skill level. Implementing a dynamic user experience will ultimately result in better ratings and reviews, helping your app stand out in the app store, leading to lift in average session length, an increase in daily active users, and growth in the number of sessions per day.
Privacy and Security
7) The first experience can be magical or "creepy", How do you ensure that the experience is not "creepy" and they keep coming back for more?
When building an app, one of the best things you can do for your users is to make the experience easier. Many times this is achieved by reducing the steps they need to do to be able to get what they want. Saving clicks and gestures is saving time. Giving a user what they want quickly gives them time back or gives you another chance to engage with them.
Being able to deliver this simple and vital experience means you have to know a lot about what your user does. Collecting, analyzing and using contextual information like sensor data, third party data, location behavior and social content are the keys to this reduction in friction.
Instead of asking for location permission right at the time of download, consider waiting until the user clicks on a feature that requires location. At this point it will be more clear to them why turning on location is useful and they will already be more acquainted with your app so they might be more likely to grant you permission to use their location.