You think you can, you think you can, you think you can!
Some perceive Facebook as the greatest persuasion platform since the tree stump. A billion Other Guys willingly (!!!) giving up personal information, sharing more information every instant, and spending more time looking at Facebook screens than any other. Now add Big Analytics to Big Data. Bang!
Facebook is delivering targeted Big Data with Big Analytics advertising to the wrong Other Guys and wrong in every sense of the term.
The case offers a glimpse into how young Facebook users are sometimes exposed to ads inappropriate for them. A 14-year-old girl in Washington state said she “liked” an ad that led to the Facebook page of a nude webcam-modeling site. A 17-year-old boy in an Oakland, Calif., neighborhood beset by gun violence repeatedly saw an ad for a concealed-carry handgun holster.
If you use Facebook regularly or read about it a lot or just occasionally read the Persuasion Blog, you know that Facebook is always making impossible persuasion errors like this with its greatest persuasion platform since the tree stump. While selling itself as Persuasion 2.0 Facebook consistently produces the wrong Box with the wrong Play on the wrong Other Guy aimed at the wrong TACT.
Given the technology of Big Data and Big Analytics, this should be impossible. Facebook owns simple, obvious, and unambiguous data and it cannot master either the data or the code to prevent these recurring stupid mistakes. Permitting children into sexually explicit websites is illegal and neither Facebook nor the Big Persuasion Muggles can figure out how to check a user’s profile for the user’s age to qualify it as legal or illegal.
The mistake here is not Facebook’s, but yours. Persuasion 2.0 is sand and ice. Big Data with Big Analytics is sand and ice. You are a Saud or an Aleut standing in a Bolivian Bank buying sand and ice. BDBA delivers no new persuasion and indeed delivers more persuasion problems than the persuasion from Your Father’s Oldsmobile. Detail the technical problems.
1. Big Data collection is crammed with errors. That firehose of your personal information static and dynamic is not pure.
2. Big Databases are not carefully cleaned for errors.
3. Multiple error-ridden Big Databases are combined which permits weird, unpredictable, and uncontrollable interactions to occur.
4. Big Analytics operates with statistical significance and over-modeling which creates prediction models that are inaccurate, data-specific, true for the once in a lifetime that has already happened.
5. Persuasion 2.0 always rides on the back of proven Persuasion 1.0 so when successes do occur, the results are confounded. How much was 2.0 and how much was 1.0? No one questions success, so 2.0 get the credit it does not earn.
Pause with this beating. Reflect.
Facebook is 10 years old. Big Data and Big Analytics are 10 years old. These are mature systems. They are not operating with beta programs. These systems have been active in the field for years and they are still making these kind of stupid, destructive, and preventable errors. Until the operators grind out high quality data collection and high quality data bases and combine that with scientific data analysis, not Tooth Fairy Tale science, you will never have useful Persuasion 2.0. And, it’s unlikely this will occur anytime soon. Back to the beating.
6. Persuasion 2.0 sells better than Persuasion 1.0 not because it is scientifically better but because it is persuasive. I can sell you the New New Thing more easily than the Old Thing.
In research, during data collection you carefully monitor how the data is coming in and look for problems in the collection and entry process. After data collection, the first thing you do is hand check each data point, often twice with different people. Did you collect what you wanted to collect and is it correctly entered in the correct location? Data integrity is a boring, tedious, and essential element of good research.
Big Data daunts this precision precisely because these stupendous batches of bits cannot be checked by humans. This leads to automatic error-checking programs that scan random (please, please, please) samples of Big Databases for predefined types of errors. If the error-rate is within an acceptable range (can we still sell this without getting sued?) then you’ve got Big Data. You cannot use an automated error-checking program to accurately check an automated data collection program. You will miss errors and enter them into your work.
Facebook as persuasion is smoke and mirrors. Facebook as social media organizing mass WATtapping is a functional network of Apps ‘n iGizmos. Networks themselves are not persuasive; they are networks. Panthers, vampires, and werewolves use networks to execute persuasion.
Again, I admire the guys running Facebook for persuading a lot of Other Guys to value the company over $100 billion. That’s an amazing accomplishment for what is only a lot of tin cans and strings. While you cannot fool all the Other Guys all the time, you can fool them enough to become a billionaire . . . which with all this peace and prosperity is not such a big deal anymore.
P.S. Another take on the Little Engine from the earlier, funny period of Saturday Night Live with Mr. Mike telling bedtime tales.
Okay, now.. “One time, there was a little train who had to pull a giant load of scrap metal up the mountain. He had never pulled such a heavy load in his life, and so when he left the valley, his little wheels said, ‘I hope I can. I hope I can. I hope I can. I hope I can.’ But, before long, he picked up speed and his little wheels said, ‘I think I can. I think I can. I think I can. I think I can.’ Soon, the little train was whizzing right up the mountain, and now the wheels said, ‘I know I can! I know I can! I know I can! I know I can! Heart attack! Heart attack! Heart attack! Heart attack! Oh, my God, the pain! Oh, my God, the pain! Oh, my God, the pain! I left my pills in the roundhouse!! I left my pills in the roundhouse!!‘ And he died.