You are here
Google Wi-Spy Was an Intentional Plan to Beat Skyhook Wireless
Submitted by Scott Cleland on Tue, 2010-11-09 17:35
Google's 'Wi-Spy' vacuuming of all of everyone's WiFi signals was no "mistake" -- as Google has repeatedly asserted -- but part of a purposeful and comprehensive Google business expansion plan to enter, catch up and compete with SkyHook Wireless, Google's only significant competitor in mobile location services. (In September, Skyhook sued Google for deceptive and unfair trade practices and patent infringement.)
- The mounting evidence suggests that the FTC prematurely dropped its Google Wi-Spy investigation based on weak and unaccountable assurances from Google.
- State investigators criticized the FTC's reasons for closing its probe, and multiple countries are continuing their Google Wi-Spy investigations, because they are unwilling to ignore Google's misrepresentations and the evidence -- and look the other way like the FTC.
- A House investigation of both the FTC and Google's handling of this Wi-Spy matter is likely given that the Joe Barton, the Republican Head of the House Privacy Caucus, just told C-Span that the Google Wi-Spy problem is "very troubling" and "...there appears to have been a conscious effort to collect information. Google said it was inadvertent, but it wasn't just kind of accidentally gathered..."
So why was Google Wi-Spy intentional?
First, the competitive impetus for Google to systematically collect all available wifi signals around the world is to improve the accuracy of Google-Android's mobile-location-services.
- The more accurately and instantaneously a provider can pinpoint a mobile user's location, from hundreds or tens of meters to meters, the more valuable advertising and referrals are, and the more advanced a provider's supplemental products and services can be.
- Simply, an advertiser is willing to pay a lot more to advertise to a specific user, if they are confident that they know how close the user is to a given place, product or service that the advertiser wants to sell.
- So what's the big deal with mapping WiFi locations? Why not just use GPS or cell tower triangulation?
- One reason is that GPS does not work indoors or in areas with lots of buildings, and GPS is often much to slow to satisfy normal user response-time expectations.
- Cellular triangulation is also not as reliably accurate enough for many advanced mobile-location-services applications.
- Mapping the literally hundreds of millions of stationary WiFi signals around the world geographically, effectively fills in the coverage and accuracy gaps of GPS and cellular triangulation. And then when the three technologies are all optimized together via highly-sophisticated software algorithms, they can provide far superior accuracy and response time collectively than any of the technologies can deliver alone.
- Long term, the ability for Google to most accurately know Android users' near exact location at any time, is worth literally billions of dollars in advertising to Google and its advertisers.
- This location accuracy capability is a big reason Google is confident that Android will become a $10b business, over ten times bigger than it is now.
- Interestingly, Google's StreetView effort was apparently Google's public stalking horse, misdirection, and cover story for Google's secret three-year scheme to collect all of everyone's WiFi signals, because mobile-location-services are orders of magnitude more lucrative and monetizable to Google than the StreetView application ever could be.
- Claiming that the collection of some of the WiFi data was a "mistake" belied and cleverly belittled how strategically valuable amassing a worldwide database of WiFi locations is to Google's expansion plans into mobile and local advertising.
Second, Google has long been way behind the market leader in mobile-location-services, Skyhook Wireless, both quantitatively and qualitatively, and was desperate to catch up. SkyHook is the only entity with 250 million mapped WiFi access points, patented algorithms, and a critical mass of data from users to build predictive behavior models.
- Google trialed Skyhook Wireless' service from 2005-2007 to learn as much as they could about their main competitor, and then from 2007 to 2010, Google, secretly and as fast as possible, mapped streets in 33 countries with its StreetView/Wi-Spy vehicles.
- It is my view, that if Google was upfront and forthright with authorities three years ago that they were also war-driving WiFi signals, in addition to taking the StreetView pictures, authorities in many Nations would have objected and been able to protect their citizens' privacy from Google's secret Wi-Spy program.
Third, Google's aggressive collection of all WiFi signals for a given location, including payload data, was a clear attempt to collect more data than Skyhook, in order to somehow figure out how Google could produce a more accurate mobile-location-service than Skyhook.
- Google's engineers are famous for having an insatiable appetite for information/data, because in the Google engineering culture the availability of massive amounts of data at Google enables engineers to measure almost anything against actual real world data.
- Simply, Google had a powerful competitive motive to vacuum up and keep all available WiFi signal data, because a prime source of Google's comparative/monopoly advantage, is its capability to analyze and monetize enormous data-sets.
Fourth, Google had to have made a conscious engineering decision to collect all WiFi signals because that decision had major design and operation implications for Google's multi-national, multi-year effort -- that belie Google's misrepresentation that this was just a simple "mistake" of "an engineer" back in 2006.
- Simply, Google encountered a major decision fork in its engineering design of its WiFi signal collection system that had dramatic implications for:
- How fast and efficient the vehicles could operate and cover territory;
- How the antennaes were tuned algorithmically;
- How the databases were organized to collect data; and
- How much storage capacity would be needed for each type of collected data.
- Google basically had a choice of two different WiFi signal collection approaches -- active (targeted) or passive (comprehensive).
- Google could have chosen Skyhook Wireless' "active" approach to protect users privacy and identity by only collecting signals on the limited channels that confirm a device was at a particular location.
- In addition to ensuring that the collection could not collect private information, an active approach only "pings" a device for a response, and only records the specific channel where a device responds that it is in range.
- This bare bones ping-signal only takes about one tenth of a second to complete.
- It's clear that Google's well-known engineering bias for more data than less, and Google's competitive imperative to create a more accurate mobile-location-service than Skyhook, compelled Google to consciously choose the more comprehensive passive data collection technique, collecting data from all available channels, which is more akin to a police scanner that is tuned to collect all signals from every channel emanating from any location.
- Importantly, this passive vacuuming approach of all WiFi signals from a location would take a few to several seconds to collect versus the tenths of seconds that an active ping-only approach would entail.
- Another big reason this was not inadverdent, is that Google had to know that what it was collecting, because it had to organize its database to collect the data and put it in the correct field, and they had to design the database capacity to store vastly more data than the data light amount that Skyhook's active/targeted or ping-only approach entailed.
- It the equivalent of having to know whether you need to have a suitcase or a cargo container available to house the expected data.
- Thus, it does not pass the laugh test that this fundamental and ongoing software design trajectory was a "mistake" or inadverdent, because it had such lasting and compounding implications for everything that followed.
In sum, Google's vacuum approach of collecting all available signals from any given WiFi location was intentional, because there was no conscious effort to filter out and not collect and store private payload data that every WiFi engineer on the planet should understand has serious privacy implications. The evidence is overwhelming evidence that this was not simply a "mistake," by one engineer at Google.
Now that Google has claimed to have stopped its WiFi signal collecting, the BIG open question is whether Google is secretly tracking its Android users in order to map these WiFi locations for Google without their knowledge? It certainly would fit with their past practice and enable them to continue to improve their competitively inferior mobile-location-service.
It is disturbing that the FTC so quickly, and completely shut down its investigation of Google WiSpy, when there are so many critical and troubling questions that have not been answered transparently and truthfully by Google.
- It raises the question of why is the FTC apparently more interested in protecting Google's privacy, than the privacy of American consumers?
- What is Google still hiding?