Data Mine All Mine

After a nice long day of doing whatever it is I do, household asleep, me very suprisingly not, I climbed into my easy chair and decided to peruse Netflix for anything haven't seen, or at least, haven't seen enough of.

The categories have changed up a bit since last time I looked.

No more is it just a list a genres, now it's got categories like "What's Hot on Netflix RIGHT NOW", "What We're pretty sure you're gonna like." And "What's Hot on Netflix Right Now In YOUR AREA!"

There is a clear descrepency between the things I like and the things my neighbors are watching, which doesn't surprise me, but the specificity of it unerves me a bit.

It shouldn't.

Netflix has my area code.

If you belong . . . then they have your area code too.

And if you're in my area code, then there's a good chance that what you've been watching is going to show up on my queue.

Yes . . . I know you've been watching Transformers Movies, though for the life of me I can't figure out why.

Actually I know why . . . cause it's the same reason I've watched them.

With a cursory curiosity that begins with "Can it really be that bad?"

Followed by the conclusion ten minutes later that "Yes . . . it really is that bad."

I admit it . . . I've watched the first ten minutes of all of them.

But never much more than that.

I swear.

Which kinda leads me to my first thought about the whole data mining thing. See . . . I gravitate towards stand-up comedy, in depth documentaries, and sci-fi movies from the eighties, so when you open up my "recommended" list, you're gonna see . . . you guessed it . . . comedy, documentary, and sci-fi movies.

Their algorithm doesn't have to do a lot of heavy lifting in my case.

It seriously doesn't know what to do with all my viewings of The West Wing.

Anyway, I saw a stand-up comedian that I hadn't seen before, and the title looked at least a little intriguing, four out of five stars and such, so I put it up and leaned way back.

I won't tell you who it is, because it wasn't funny. Not a dismal trainwreck, but not funny. Set-ups that were tired and punchlines with little poetry.

It got my ten minutes.

But that was all.

I ended up settling in for a documentary on food labeling.

Did you know that sugar is bad for you?

Of course you did. But . . . did you also know that the recommended daily allowance is 25 grams, which is so far below the amount that is in just about every serving of processed food, that food companies don't list the RDA or the RDA percentage on packages?

Like if you were to eat four fig newtons, you'd be done for the day.

Food companies don't have to tell you that because technically, sugar isn't a nutrient.

They'll tell you the grams but not how it relates to a healthy dose.

Interesting . . . at least I found it interesting . . . which is a lot more than I can say about the stand-up comedian.

But here's where I get back to an earlier thought . . . see . . . it occured to me that Netflix knows how long I watched the routine.

It knows I got about ten minutes in, gave up, and went somewhere else.

And if it knows . . . is it possible the comedian knows?

Not my viewing habits personally, but do you think he gets a report that shows exactly how long a percentage of the watching population got before they got bored and went somewhere else?

Wouldn't it be possible that Netflix could dial up the exact joke that made people go "Ugh" and then send the comedian an email that says something like "You should probably get rid of that airline food bit . . . it's really not working."

Like I get a report each month that tells me how many songs I've streamed, which titles were more popular than others, but what I don't get, which might be nice, is how long the song was listened to.

Pandora, could, but they won't, tell me what percentages of my songs were skipped after the first thirty seconds to a minute.

They can, and they do, know when I've been skipped, and they adjust the number of my songs in the rotation accordingly. But I don't get any of that. I just get the percentage of a percentage of a penny.

And then I got to Taco Bell.

Which I can't afford.

Facebook does a little better job with sharing some statistics like that.

Like I know exactly how many sets of eyes scroll past my posts each day. Of those sets of eyes, I know exactly how many people clicked my link.

That's something I can use.

See . . . if fifty people see my post . . . and five actually click to read . . . that's a ten percent viewership.

Now let's say I wanna boost that on a friday afternoon, and I talk about pet ownership with a picture of a cute little cat and I get 100 views with 13 clicked links, I know that cute cat pictures can increase my readership by 3%. Not bad, but there are only so many cat jokes I can make before I start offending people.

Just some fun facts: I get around 10% when I use pictures, 7% when I don't. That's seems really obvious, but it's nice to have the numbers.

It also tells me my demographics. Did you know if you're reading this you are most likely a female aged 25 to 55, married with children, and you lean slightly to the right politically.

That means the family stuff works like a charm, but Donald Trump jokes do about as well as Hillary Clinton jokes.

That's what your data tells me anyway. Or . . . more specifically . . . that's what Facebook tells me your data tells me.

It could lie.

How would I know?

Most of you have some college. Thanks for that. Does my heart good.

But that's about it. I don't know who you are or where you live or how many fig newtons you snacked on, but I'm willing to bet it was too many.

Which leads to kind of this classic conundrum. See . . . the thought of data mining is a little scary.

A company used to have things like focus groups of ten, maybe twelve people pulled from wandering the mall. Now they literally have multi-billions of little decisions made by us everyday so they can tailor your purchasing to your statistical demographic.

It's not new. And in the case of helping a fledgling comedian amputate lame material, it could actually make the things we want, even better.

Like Guitar Center knows me pretty well. They know I come in every few months for strings, so I always get the email about string sales, and they know every few years I come in for a big ticket item, so I always get their big beautiful full color brochures.

They don't know I will never buy another pair of Martin acoustic strings again, no matter how good the sale is. Those things sounded like crap and broke on me during a really important show and I'm very unforgiving about things like that.

So the system isn't perfect.

But it can be a little creepy. Like Target knows when you're getting married, when you're moving, and knows you're pregnant before you tell your family. There isn't a lot about you that can't be gleaned with a thorough search of your browser history . . . so if you're a private person . . . time to get off the grid.

I'm not an alarmist. Though I can see why anyone would be. Like, the FBI showed up at this guy's door one day because he'd been searching for three days through hundreds of websites about poison, wife killing, hiding a body, and removing DNA evidence.

He wasn't a killer.

His wife was very much alive.

He was a script writer for CSI.

Doing research.

You could see how that could be confusing. And disconcerting. To both parties.

It's going to be interesting to see what happens when the "Too Much Information" Generation starts becoming political/economic leaders. Like . . . thank god we don't have twenty years of Tumblr posts from Donald Trump to have to go over. Could you imagine?

There's a Hillary thing with emails going on, but, f@#$, I so don't care. But imagine if she were forty years younger . . . would we be subjected to information about nude selfies and Tindr hookups?

Will we in the future?

That's very much possible and frightens me a lot more than the Thought Police.

Thankfully, we still have a good chunk of the first amendment alive . . . there's still some built in anonymity to the internet . . . and a good rule of thumb regarding anything you do online: Pretend like your mother is reading . . . cause she probably is.

And don't eat so much sugar.

It's really bad for you.  

No comments:

Post a Comment