By Andy Gordon on May 21, 2014 Dear Early Investor,
Have you noticed? It's a funny world we live in...
Where planes disappear and the weather still manages to surprise.
Yet the moneymaking success of movies is known not days in advance of their openings, but months.
And your personal habits and needs are increasingly becoming an open book to anyone willing to pay for the privilege. Retailers can know when you're going to buy your next lawn mower or car. Not only that, they know if you're going to want gas-powered, plug-in or battery; how much you're willing to pay; and what features you're willing to pay extra for.
Peering into the future has a bit of a history.
For the last few thousand years, up until a few years ago, predictions were the exclusive province of prophets and madmen. From the oracles of ancient Greece to our modern wizards and witches, it has been among the most prized of skills.
In 560 B.C., Croesus, king of Lydia, wanted to see which oracle gave the most accurate prophesy. So he sent out emissaries to seven sites. They asked the oracles the same question: What was the king doing at that very moment? Pythia, priestess to Apollo at Delphi, was proclaimed the most accurate oracle. She correctly reported that the king was making a lamb-and-tortoise stew.
Since oracles were believed to be the portals through which the gods spoke, they were never wrong. Any inconsistencies between what was predicted and what transpired were dismissed as a failure to correctly interpret their responses.
And this, believe it or not, is still happening today. A few thousand years later, people remain the biggest obstacle to a world of increasing predictability.
The Rise of the Corporate Oracles
The oracles haven't uttered anything of consequence since A.D. 362. Predictions now come from "big data." And the gods who can see the future don't live on Mt. Olympus. They're the companies that track and analyze big data.
Like Google. Its data comes from the minutiae of our lives. So it has to be massive.
Everybody knows about Google. But Google is not big data. And big data is not Google.
Google may be leading the charge. But not for long. And perhaps not anymore.
A startup called Networked Insights has out-Googled Google. It "reads" more than 560 million posts on Twitter, YouTube, Facebook, Tumblr and other sites... every day.
These numbers are impressive. But where the company really excels is its reading ability. It knows the difference between Bear and bear, for example. Google itself says that reading ability is key. It's developing technology that will "read and understand" via artificial intelligence. But it has a long way to go.
In the meantime, Networked Insights seems to have jumped ahead.
It can predict opening weekend box office within $5 million (or 78% accuracy). Google is less accurate. More telling, Google does it one week out. Networked Insights predicts it more than three months in advance.
From Predicting the Fall of Empires to the Flaws in Smartphones
The power to predict behavior is mostly focused on consumer behavior. And for obvious reasons. Retailers are willing to spend big bucks for such knowledge. And they're getting their money's worth. Consider... - Networked Insights found out that both Android loyalists and iPhone "fence-sitters" valued screen size and price. Samsung used this information to steal market share from Apple.
- 6Sense can track people browsing your product webpages. It can track them downloading your product information from other sites. And it knows who has purchased accessories to your product. Cisco says that 6Sense's product helps it predict when accounts are about to close more than 80% of the time. Cisco's conversions have increased tenfold, as a result.
- Another startup, Estimize, crowdsources earnings estimates for stocks. About 3,000 analysts and 15,000 community members provide the data. It says its predictions are more accurate than Wall Street's 69% of the time.
The fly in the ointment? Human error and the old "garbage in, garbage out" conundrum. More than 2,000 IT leaders were surveyed last year. Sixty percent believed their organizations lacked accountability for data quality. More than 50% questioned the validity of their data.
But we've seen accuracy rates of 69%, 78% and 80% achieved in the above examples. And the data will get bigger and better, if only because the technology will increasingly be able to pick out anomalies.
Today, buying behavior is getting the most attention. But big data is spreading. It's already fomenting changes in how election campaigns are run (read how Obama won the election here)... how movies are rolled out... and how we invest in stocks.
It's only a matter of time before big data invades the startup space. Imagine what happens the second a company announces an IPO...
Technology will be listening in on the conversations of bankers and brokers relating to the event. Crowdsourcing systems will be eavesdropping on nonprofessional investors.
The technology already exists to do this.
We'll soon see the day when price movements of new IPO shares will be predicted months in advance. Seasonal earnings events will also go through the same kind of metamorphosis.
The Las Vegas in the Sky
Ultimately, all our social behavior will be subject to predictive analysis. Even the unpredictable "Black Swan" events will become susceptible to predictive algorithms (MIT is hacking away on this problem right now).
We're headed toward a world where almost everything we do is measured by powerful digital oddsmakers watching down on us from the Cloud (not unlike the gods of ancient Greece keeping an eye on us from above).
Big brother doesn't look like a giant, all-controlling government. It looks more like a highly accurate Las Vegas minus all the fun.
There's something deeply disturbing about this.
It's all too neat. People don't like being put in a box. There will be movements to opt out of the Cloud. A subculture will emerge, made up of behavioral feints and digital misdirections, meant to lead predictive technology astray.
A new science will be born on the behavioral consequences of predicting behavior. In my IPO example, potential investors will be focusing less on the IPO itself and more on the reaction to the IPO price predictions.
I think in numerous spheres of human behavior, including investing, people will defy predictive analysis to a surprising extent and remain largely inscrutable.
I, for one, don't want to develop the technology that's going to tell me this early who's going to win the next presidential election or, for that matter, win the World Series.
And I don't think I'm alone.    
Recent Articles From Early Investing By Andrew Gordon on May 16, 2014 Lyft, if you don't know, is Uber's main competition. Uber is the startup primarily responsible for blowing up the taxi business. Both companies let people call up rides from their smartphones. They're at war right now...
By Andrew Gordon on May 13, 2014 Can Washington get crowdfunding right this time? Only if it begins to worry about everyday investors as little as it seems to worry about our country's wealthy investors...
By Andrew Gordon on May 9, 2014 The current startup scene represents a different mix of opportunities and challenges. Knowing what I know now, I'd put together a bigger and more impressive team serving two purposes. One: Increase your chances of success. And two: Win the trust of potential investors. |
No comments:
Post a Comment
Keep a civil tongue.