Monday, April 27, 2015

Mobile Monday's Geospatial Big Data In Silicon Valley

I trekked down to Silicon Valley last week for my regular taste of Mobile Monday.  The Silicon Valley campus playing host to this particular event had one of these post-modern water sculptures out front.  I see these things so frequently now on such "campuses" that I'm pretty sure the big tech firms are trying to subtly outdo each other with understated water installations.  The drought is still on in California but these displays all claim to use recycled water.  Okay, whatever.  I did not take any photos of the water display or its sponsor.  You'll just have to believe me when I say I was there.

Anyway, geospatial is shaping up to be another next big thing now that all the other next big things - Web 2.0. cleantech, social/mobile/local - have run their course.  The leading geospatial player on my radar is DigitalGlobe.  You may have seen their work cited in open sources when US military officials with NATO used DigitalGlobe photos to bolster their argument that Russia was using military force in Ukraine.  The DigitalGlobe rep who spoke at Mobile Monday made a clear case for mining geo-linked data sets.  Making those data sets available to retail users in real time will take a lot of bandwidth.  Fortunately for DigitalGlobe, plenty of users love playing with high-resolution maps.

The experts on hand discussed geotagging as a user engagement strategy.  The good news for them is that incentivizing users to tag images is easy with some gamification experience.  Users who score can unlock "expert geoanalyst" badges and build their reputations in open-source imagery analysis.  Geodata startups should pay attention to exploiting all the free labor they can get in finding a mass audience for their analytical solutions.  The best freelance analysts will eventually demand to be paid premiums, much like programmers who become repeat hackathon winners.  You heard it here first at Alfidi Capital.

I am not aware of any accelerators specifically focused on geospatial startups.  I expect that to change as companies like DigitalGlobe succeed in monetizing crowdsourced geodata.  One of the expert panelists mentioned how years of map data add context to whatever users do with a download.  I would add that years of embedded links from news articles and social media feeds can add more searchable context if the download sets were amenable to enterprise knowledge management solutions.  The difference between layering and filtering data matters little to retail users but becomes more salient for knowledge managers farther up in a large enterprise.

It's time for some personal stories that add color to the geospatial sector.  My own experience with geotagged image data dates to 1996, when I was on active duty in the US Army.  In the '90s I worked with systems that used scanned 2D maps overlaid with crude geotags.  The geotags did not connect to embedded data and the maps were poor simulations of 3D terrain features like elevation changes.  The military systems I worked with since 2008 showed vast improvements in both 3D rendering and embedded links.  I know from experience how enterprise search offers a compelling way for geodata to add value.  In other words, I know what right looks like.

Here come my predictions for the geospatial sector.  I expect crowdsourced geotags will be worth more if they are segmented by user competence.  Data providers should ask taggers to initially self-identify their expertise in recognizing image anomalies or data elements.  It's worth investigating to see if gamifying mass involvement will truly identify skilled analysts.  I expect data purveyors to pursue bifurcated pricing models, with one payment track for enterprises and a much cheaper track for individuals.  It will look like software pricing strategies that chase seat counts, but the winning startups will know how to cover the variable costs of processing and storage.  Geodata startup founders must read Cloudonomics if they want to win.

I would not be surprised to see the emerging relationships between geospatial sector firms and Big Data firms to lead to mergers.  The VCs chasing geospatial startups are going to be disappointed once they discover the very high costs of putting satellites into orbit.  The only possible entrepreneurial disruption available there would be from some space launch technology that does not use a traditional multistage booster to escape earth orbit.  Rocket sled launch technology would be great if it relied upon a railgun for its initial propulsion.  I respect SpaceX for getting the conversation started but I don't understand why they still seem stuck on rocket boosters as their tech mainstay.

Geospatial enterprises will be fun to watch in the next few years.  Lots of startups will jump into it thinking they have some app that DigitialGlobe or Google would love to acquire.  If said app reduces the cost and speed of processing embedded map data, they just might have a chance.  A bunch of VCs will throw money at any startup with "geo-something" in the first line of their business plan because chasing fads is in Silicon Valley's DNA.  I'll be around to laugh at the VCs who fund the worst ideas first, and to congratulate the best ones that win.

Full disclosure:  No position in DigitalGlobe (ticker DGI) at this time.