The other NY Times innovation

I say "other" because I'm not talking about the pay wall that the paper is putting up this week which has received huge attention. I noticed that the New York Times has given more prominence to its personalized recommendation feature (italics mine.) I've been a long time reader of the Times and think it is without a doubt the best newspaper out there. They have tried different content layouts through the years (Select, Reader, etc.) which have met with middling success. And now have moved to give this feature more house banner visibility on it's site.

Recently they have started suggesting articles once a reader reaches the conclusion of an individual article on their site. These suggestions take the shape of banners which slide out from the lower right hand corner of the screen. In my experience, their matching ability to what I have been reading to what might interest me has been pretty good. The suggested articles are fairly topical and relevant to my interests based on what I just read. It is a helpful feature and I find myself drifting from article to article or even rushing to the end to see what it might suggest next.

Now the paper has a compilation of these recommendations in a top 10 list on the right rail of their section front doors (eg Business, Technology, Style, etc.) although I couldn't find it on the site this afternoon. The feature does require authentication, presumably to remember my list rather than collect info about me.  In any case, perhaps they are tweaking it as only two of the ten stories suggested for me seemed relevant. Maybe this is the recommendation cold start problem since I have only read a few online stories recently or there is a kink in the predictive algorithm. In any case, I hope this feature continues as I welcome recommendations from a trustful source who values my time.  We'll see how this story develops over time as they presumably collect more info on my reading habits.

SxSW recap

Wow! South by Southwest was huge this year! This was my third time in the past 4 years. In that time, registration exceeded 3x the attendance only 4 years ago. It's been discovered. Still the conference was wonderfully nourishing (thanks to my dear friend, Delphine, for that description) for both mind and soul. 4+ days of innovation, creativity, and entreprenuerial spirit by nearly 20,000 participants.

There didn't seem to be a breakout technology this year like twitter or Foursquare/Gowalla in past years. If anything, here are the trends:

• The iPad made it's debut as a hardware platform. Tt was definitely a pro Apple crowd-PCs were in conspicuously low profile even with HP and Microsoft with concourse displays. I predict next year will have a number of sessions talking about the future of tablets, touchscreens, and this new platform.
• The entreprenerial spirit. Panelists and participants talked about life in a startup. Bizspark had a competition among several startups. Despite the rough economy (maybe becuase of it) small businesses were alive..and maybe well. At least they were surviving. And on the larger front, Groupon is thriving, facebook is huge (still), and there are many emerging and established not-so-startups waiting to go public.
• Mobile. Everyone knows this will be huge but is not sure how to implement it yet. Lots of great sessions on apps, gellocation services and small screen design. Next yeat this will be huge once we figure out mobile commerce.
• Game-ification. Making everything have some type of game. I think this is a fad but there were a number of sessions about it. No, not everyone wants to this, people! The tech community spends a lot of time navel gazing, I think,

The City of Austin was also a clear winner. The area around the convention center has been discovered by both temporary corporate marketers and permanant local businesses as a good place to set up shop. 20K people returned to their corner of the world to spread the word about how great Austin's scene/weather/people are. (Please, don't move here. Visiting is great!)

On the personalization front, there were a few panels that addressed issues around data privacy, recommendation engines, and push for relevancy. Had a stimulating conversation on a panel hosted by a rep fro Bazaar Voice and Zaaz on exchanging value for privacy and what that means to the futgure of e-commerce.

Thanks to the organizers, volunteers and participants who keep this such an energized and valuable confernece. I highly recommend everyone attend...well its seems like most already do. Soon it will be so crowded, nobody goes there any more.

Verizon is slated to sell it's new version of the iphone next week after keeping the world waiting for 4 years. Analysts predict a stampede to sign up. Initial reviews report that the call quality, coverage and number of dropped calls (very rarely) is impressive. Yes, it is probably a better calling experience. But I think are a number of factors at work that will make the stampede non-event:

• Most of the population is already locked into a 2 year contract of some kind. With Verizon or other carriers. The penalty for ending a contract early is expensive ($300+) as is buying an unlocked iphone. As contracts end on a rolling monthly basis, there maybe a long term trend to move to Verizon but  not an avalanche.
• The iphone 5 (or whatever it will be called) is likely to be released this summer. One can't take their AT&T iphone and put in a Verizon SIM card to activate. Verizon's iphone is different hardware. Who wants to buy a a new Verizon iphone and have it be old technology in 5 months?
• While Verizon's call quality is superior, it's customer service is not. Neither is AT&T's but there is downside to switching. And the rates offered are almost exactly the same.
• Will the iphone be enough to get people to switch to another carrier? Android is gaining fast with many more models to choose from. I love the iOS but of the 3 iphones I've owned, none has made it through the 2 year contract without a major hardware or software malfunction that makes some features inoperable.  Inertia is powerful. 

So, we'll see. I think there will be a gradual shift to Verizon (5% market share) but not the stampede many have predicted. Which is too bad, actually. It would make a great business school case on how a carrier is pulled by the device which then switches once a choice is available in the marketplace. 

Browsing without a trail...

Some ruckus this past week about the how the new editions of Chrome and Firefox will include the ability to surf without a trace. No cookies, no clickstream, no IP address. No one would see you come and no one would know after you were gone. Or track you later. Sites would have to ask permission to identify you at entry, follow you onsite and then track your departure. Consumers think this sounds great.  Most don't know that online companies do this (and much more) today already. But governments do (or their younger aides do) and regulation is an election year away.

Most consumers blindly surf away today, 'friend' this, or 'like' that with reckless abandon. Almost 'surfing out loud' for all those to see. How many have Facebook 'friends' they have never met? I bet most do. And how many know that they are monitored throughout their site experience by savvy online marketers. If their site experience is rewarding, most probably don't care. They are getting something in return--a more relevant experience. The site has developed trust with the consumer. (Great overview of the current corporate trust barometer by Edelman PR.) In an online landscape where authenticity is difficult to establish and easy to lose, people are seeing value in trustful relationships (where credibility is cultivated) and cynical when its not (being 'defriended' or following someone disingenuine) and smart sites will attract trust by delivering value in exchange for identity.

I think there is a silver lining here. My prediction is that this anonymous surfing capability will be introduced by most if not all browser companies as self-regulation.  It will probably be an opt-in to block tracking. This will already limit its usage as most consumers are slow to upgrade browsers (how many of you out there on IE7.x?) in addition to those who won't or can't figure out how to opt-out. And larger online companies will sharpen their opt-in experiences by forcing authentication or log-in and messaging their enhanced site experience for those who identify themselves. That's where personalization comes in. Ever shopped on Amazon without logging in? It's OK but not great. If the site can deliver the goods--a more personalized experience--people will trust them. They will say, "I'll let you know more about me if you reward me for it." Good sites will. Bad sites will become strictly transactional, one-size-fits-all, one and done shopping sites.  It will be haves and have nots. Gated communities and open neighborhoods.

Privacy regulation could be the best thing that happens to personalization in the near future. Let's watch and see. 

The secret sauce of Pandora

Pandora has been a trailblazer in collaborative filtering and the ability to help music fans listen and discover music combed from their Music Genome Project. Fast Company has a great article this month to give a view behind the scene of Pandora including a video interview with their Founder, Tim Westergren. Here's the video:

What I like about the video is his acknowledgement that the secret isn't the matching algorithm that powers the music suggesting service, its the people who categorize the music. And the enormous effort that it takes. The video snippets of scores of people sitting at long tables ostensibly listening and categorizing music was a little sweat shop like...albeit one of many music lovers who are selectively screened for their musical knowledge and ability to discern musical genres.

This was doubt meant to be a favorable article for Pandora, it didn't acknowledge how Apple's iPhone and itunes saved Pandora by providing a scalable revenue stream to them. Pandora rounded the corner from good idea stage to personalization stalwart when Apple signed the deal to allow Pandora listeners to download songs heard for a cut of revenue from each download. The success of the iPhone pulled Pandora along for the ride to the good fortune of all of us. 

Personalizing the Car

As noted in a previous post, I recently acquired a gently used car. It is a '08 Honda Civic Hybrid and came with their navigation system. The "Nav" as Honda likes to call it, is a screen in the middle of the dash that provides GPS service and is the display for the audio system. This the trend these days on new cars, gives the car a personality, and provides the driver and passenger a richer engagement experience. Sounds good, in theory.... As you can see from the picture, the screen is lined with identical buttons with cryptic labels. If I need to keep my eyes on the road, I better know which button does what pretty damn well. This is the least of it's problems. (Would love to play around with Ford's new Sync system. Mark my words: every car will have something like this by 2013 models.)

The GPS map service is powered off a DVD and satellite service and works competently--it displays where I am relative to my surroundings. I can zoom in and out to different scale although the map looks a little cartoonish with jagged lines for most streets. My favorite worthless setting is the far out zoon with scale of 350 miles per centimeter of screen. This places me in North America on a screen showing the western hemisphere. Not very helpful for driving around town. I've never figured out (actually taken the time) to learn how to have the car talk to me and guide me to my destination. The manual is about 1 inch thick.

The audio system's UI is a nightmare. The 'designer' of the UI must have been an engineer who has never designed a user interface, driven a car or listened to music much less done two of these three at the same time. The primary interface shows the radio station, the assigned shortcut button, and then two large touchscreen buttons to choose the background of the screen or assign the station memory preset buttons. The reset button touchscreen option seeks the local stations and then assigns them to your preset buttons that line the bottom of the screen. Great, except that you are likely to do that once. Why would you do this often if you like the stations you've selected. Stations and tastes don't change that often? But that button is always there. And if you accidentally touch it, it will reset all your stations and there is no way to stop it. A painful inconvenience to reset them all and reassign each of the six presets. The music background lets you choose a equalizer background or a watter ripple that moves to the tone of the music--sounds cooler than it is. It is a black and white screen and lacks any sense of excitement.

The point is, why was this interface allowed to launch this way?! Here's a $1000 option with a terrible user experience. The product manager for this devise should be forced to reset every owner's radio preset buttons everyday for life. And, wouldn't it be great if Honda let you choose the skin of your Nav or at least refreshed it if you bought a new DVD (about $200 at your local dealership) to right this UI injustice? If anyone knows a hack to do this, I would be forever in your debt. 

I am my Cart

Great article in the New York Times Sunday magazine My Cart, My Self about what the personalization features of Amazon, Netflix, Pandora and the like say about knowing one's self. The author laments that some of the recommendations are uncanny and some are crazy. He states "The only thing worse than being misperceived by a machine is being expertly perceived by one." 

As product recommendation engines become better and power more content, they are every-hungry for the slightest behavior or inclination that could indicate a preference, bias, or proclivity to like something. That may work great in a vacuum when it is action and response but most people's lives are a little more complicated than that. As the author explains, he shares the account with his girlfriend and manages the Netflix family account which further muddies his eclectic taste in film 

What annoys me about the recommendation “engines” that drive e-commerce nowadays is their denial that I possess free will, or at least their denial that I often exercise it. I’m told that the programs rely on algorithms that mimic the workings of my cultural intellect, which must mean that I’m an algorithm, too. I’m not, though. I’m a complicated oddball, eccentric, eclectic and erratic — prized traits now reduced to irritable rejection of ubiquitous recommendations on my e-commerce sites.

As consumers move beyond the delight of relevant recommendations and start to notice the near and far misses, how will recommendation product engines react? Amazon has a nice treatment in its account management portal to allow consumers to review past purchases and declare whether a previous item was a gift (and therefore not relevant to personal tastes) or simply not to consider the purchase in future recommendations. This is a great feature I've used to view my 13+ years of purchases (like a time capsule of retailing!) but Amazon does nothing to promote this capability to the masses. Maybe because they don't want people to try to battle the algorithm. It will be interesting to watch Amazon's decision to communicate this capability to the public and decide that consumers are ready to guide their own recommendations or continue to be a victim of it. And of course it only works on purchases and not browsed items. You'll still get the recommendations on site and through email on that item you looked at once passively or researched but had no intent on buying.