Thursday Jun 20, 2013

The Innovation Pivot

One of the things our Retail Applied Research team tries to do is "fail fast."  That doesn't mean we're trying to fail, but we want to arrive at a failure or success assessment quickly so we minimize investments in failures.  But just because a project isn't deemed a success doesn't necessarily mean its a failure.  In many cases we can pivot, reusing some of the knowledge and technology but applied in a different context.  In some circles, entrepreneurs are encouraged to run with an idea until it fails, then pivot in a different direction.  There are many famous examples of pivots, like the emergence of from a social app targeting gay men or the pivot of Tote into Pinterest.  Sometimes the original idea just didn't fit, and other times the market changed and required re-assessment.

The Austin-based start-up Digby has pivoted twice. Digby first created a mobile marketplace on the Blackberry where consumers could order products from multiple retailers in a single application.  I once used it to order flowers for my wife.  This was in the early days of smartphones when mobile commerce was in its infancy.  Then when the iPhone was released, retailers wanted their own dedicated app so Digby moved away from the marketplace app and focused on mobilizing e-commerce sites, providing mobile apps to retailers supporting iPhone, Android, Blackberry, and Windows Phone. While that was a successful business, the market was shrinking because e-commerce vendors, like Oracle ATG, started offering mobile extensions so that third-party software was no longer needed.

Wanting to leverage all their experience with mobile technology, Digby next pivoted to Localpoint, a product that enables retailers to easily create geo-fences and provide direct marketing to consumers via their mobile devices. recently gave them an honorable mention in their 7 Hot Mobile Startups to Watch in 2013.  They seem to be making great inroads with retailers using this latest approach.

Conventional wisdom says you don't know what you don't know, so its best to dive in but be prepared to pivot.  (Idea-driven vs Outcome-driven is an interesting debate and something retail labs should consider.) Agile retailers need to test (a step Ron Johnson skipped over) lots of concepts before finding the ones that work, and then not stay married to those concepts forever. And don't think this advice is limited to small companies -- large companies can pivot too.

Thursday Aug 30, 2012

5 Ways to Determine Mobile Location

In my previous post, I mentioned the importance of determining the location of a consumer using their mobile phone.  Retailers can track anonymous mobile phones to determine traffic patterns both inside and outside their stores.  And with consumers' permission, retailers can send location-aware offers to mobile phones; for example, a coupon for cereal as you walk down that aisle.  When paying with Square, your location is matched with the transaction.  So there are lots of reasons for retailers to want to know the location of their customers.  But how is it done?

I thought I'd dive a little deeper on that topic and consider the approaches to determining location.

1. Tower Triangulation

By comparing the relative signal strength from multiple antenna towers, a general location of a phone can be roughly determined to an accuracy of 200-1000 meters.  The more towers involved, the more accurate the location.

2. GPS

Using Global Positioning Satellites is more accurate than using cell towers, but it takes longer to find the satellites, it uses more battery, and it won't well indoors.  For geo-fencing applications, like those provided by Placecast and Digby, cell towers are often used to determine if the consumer is nearing a "fence" then switches to GPS to determine the actual crossing of the fence.

3. WiFi Triangulation

WiFi triangulation is usually more accurate than using towers just because there are so many more WiFi access points (i.e. radios in routers) around. The position of each WiFi AP needs to be recorded in a database and used in the calculations, which is what Skyhook has been doing since 2008.  Another advantage to this method is that works well indoors, although it usually requires additional WiFi beacons to get the accuracy down to 5-10 meters.  Companies like ZuluTime, Aisle411, and PointInside have been perfecting this approach for retailers like Meijer, Walgreens, and HomeDepot.

Keep in mind that a mobile phone doesn't have to connect to the WiFi network in order for it to be located.  The WiFi radio in the phone only needs to be on.  Even when not connected, WiFi radios talk to each other to prepare for a possible connection.

4. Hybrid Approaches

Naturally the most accurate approach is to combine the approaches described above.  The more available data points, the greater the accuracy.  Companies like ShopKick like to add in acoustic triangulation using the phone's microphone, and NearBuy can use video analytics to increase accuracy.

5. Magnetic Fields

The latest approach, and this one is really new, takes a page from the animal kingdom.  As you've probably learned from guys like Marlin Perkins, some animals use the Earth's magnetic fields to navigate.  By recording magnetic variations within a store, then matching those readings with ones from a consumer's phone, location can be accurately determined.  At least that's the approach IndoorAtlas is taking, and the science seems to bear out.  It works well indoors, and doesn't require retailers to purchase any additional hardware.  Keep an eye on this one.


News and ideas about the retail industry with a focus on customers, innovation, trends and emerging technologies.

Oracle Industry Connect 2016

Stay Connect with Oracle Retail


« June 2016