Data Mine

Capsule:  If a tree falls in the forest, who owns the data that records when and where it fell?  No one knows--there are too many variables--and major fear factors are discouraging media companies from making the claims necessary to build new data-related products.  Concerns about the perils of media data mining are holding back a potent video evolution, from our wired TV systems to the internet.  It's time to move forward, before advertising challenges wreak more havoc and take more money out of the media. (www.mediapost.com/publications/?fa=Articles.showArticle&art_aid=113863)

Whole industries were taken out and recast in the economic collapse of 2007-2009, including financial services, the auto business and, to a smaller degree, the advertising business.  Financial services are looking healthier, but they're still on a potent Fed drip, as is the auto industry.  Like doctors attending to seriously injured patients on morphine, the government is trying to find the best way to help the banks and the car-makers heal, after which they'll disconnect (we hope) and American industry and consumers will weather the withdrawal symptoms that will surely follow. 

Meantime, no one is rushing in to medicate the media.  Advertising has also crashed and has taken some of America's proudest media brands on a non-stop thrill-ride without safety belts, throwing a few newspapers from the train and leaving many other media properties, including two out of the three major TV broadcast networks, looking green--and, not in a good way.

If anyone had told us in the balmy days of 2006 what was ahead, we wouldn't have bought it--even though we were apparently buying everying else.  What ensued were two straight years of global economic disaster.  Are we done yet?  We'll see. 

These were also the years of the NebuAd and Phorm controversies--one, here in the US, and the other, in the UK.  Both NebuAd and Phorm were and are start-up software and network management systems built to place advertising online for wired customers targeted to advertisers' specific interests based on consumers' online surfing behaviors.   Congress and the FCC threatened to investigate NebuAd's privacy architecture, following a west-coast cable broadband test.  US citizen activists protested in advance and the cable company testing NebuAd backed away for now. 

In the UK, Ofcom, the official government media regulator, initiated a major review that determined that Phorm's "anonymized" data mining should be permitted; but, like their US brethren, UK-based privacy activists were vocal and persistent in descrying commercial use of surfing data.  Neither targeted online advertising effort has progressed in the US or the EU.

All the while, cable companies have been trying to figure out how to mine and use their TV viewing and broadband data.  The arrival of digital TV, complete with a significant share of two-way set-top box real estate, makes it possible for cable to use information about its customers' viewing preferences in the ordinary course of business, as long as that data is encrypted, not personally identifiable, secure from mischievous hackers and employees and not sold to third parties.  If a cable company is interested in selling data for any reason or in reviewing data in anything other than an anonymous form in the ordinary course of business, it must ask its customers to opt-in to the process regularly and explain what its data mining process will be as well as its target yield.

Satellite companies have been progressive in some ways with data collection through a small percentage of their set-top box subscriber base, but have not expended a great deal of capital on either this or their virtual two-way advertising capabilities, despite some impressive pilot projects.  The telcos have built two different very advanced data collection products--one that operates just like a cable system and one that operates like a complex amalgam of online services that we all know by now follows us everywhere, stores everything, and then sits paralyzed trying to figure out what to do with a now open data mine. 

And Tivo, always ahead of the crowd but short on cash--the quintessential American company, as it turns out--has a method and may have the beginnings of a cure to the data mining madness.  Like Amazon, Tivo has long stored and used its customers' viewing data, playing back suggested viewing choices based on what customers seem to like.  In other words, Tivo watches what you watch in order to help you get a better TV experience.  Shouldn't the cable, satellite and telecom industries be doing the same thing?

The ultimate TV GPS will help customers understand what's on their wired systems and what can complement those viewing choices on the internet from local, national and global sources.  In order to bring TV and internet viewers to their destination--better TV?--TV GPS will have to start by fixing on where its viewers are.  In the mobile world, that means literally identifying the physical location of a cell phone.  In the wired world, fixing on a viewer's starting point will start at home.

Wired systems already have a plethora of data sources identifying the media habits inside a house.  Pegging people inside the house is a different challenge, but our home sweet homes and all of the modems inside, including those inside of digital set-top boxes, are well known.  Some cable distributors have begun to update their data operations with linked data on cable modems and set-top boxes that can actually be "read" in order to enable customers to move video content from their PC screen to their TV screen and back.  The TV Everywhere concept will push most of the cable (and therefore the telco TV) world in this direction.

Why can't this data be mined for better navigation?  If it's possible to maintain, secure and encrypt the data bases that power suggestion engines, it should be possible to use mined data without falling too far afield of regulatory intention when the customer is doing the mining.   Cable companies have a propitious opportunity that could be a game-changer.  They also have a logical interest in offering navigation that, like Tivo, takes you to familiar places based on where you've been for a fee.  Think of cable TV GPS like a for-real TV Guide.  You pay the guide or, at least, tip him; why wouldn't you pay your distribution company for making sense of the media wilderness?

Up until this point, cable bandwidth concerns and the public equity markets' over-reliance on evaluating the cable industry according to its ability to reduce its natural (and winning) capital-intensive spending have held true TV GPS at bay.  Polling two-way set-top boxes for viewing data that can then be mined and played back as viewing suggestions is an elaborate and network-intensive enterprise.  It costs a lot of time and money to do on the cable platform itself. 

But, there are ways to combine network-based navigation and targeted promotion on a phased basis without breaking the cable bank.  All of the cable distribution majors have some version of a network-based navigation system in the works and in the field.  Of these, there isn't one yet that isn't design-constrained by a fear of digital rights management lawsuits, lost advertising monetization opportunities or inflamed federal regulators.  In other words, no one is designing TV Google...yet.

While big distribution (including alternative distribution companies like navigator Tivo) and big content keep their eyes and wallets on the game of "this is mine and that's yours until it can be mine too," some company or distribution-related industry will draw a simple conclusion: as long as the customer owns their own data and pays for a service that helps them use it to learn more and watch more and buy more, there are fewer risks in data mining than may meet the eye. 

The far worse risk of watching and waiting rather than improving content and advertising could precipitate another media storm enabled by the lack of a floor under TV advertising rates and its likely next effects on big content.  The cable content players have seen advertising revenue improve at the expense of their broadcast predecessors.  Over-confidence is unlikely to be rewarded anytime soon.   Media would be wise to get recession-proofed with new products that support its economic evolution as we all wait for the end of the recession itself.