The Problem With a Billion Tweets

“It’s not what you know, it’s how you act on what you know,” William told me.

William was William Donald Schaefer, the governor of Maryland and my boss. He meant it’s not important to know every last detail of an assignment or task. You can always fill in the gaps later.

When I started to invest in businesses, I adopted a different approach. Taking advantage of opportunities was no longer about moving fast but moving smartly. So I took up Buffett’s motto: “Risk comes from not knowing what you’re doing.”

It meant acting more deliberately and making fewer mistakes, even if it meant some fast-moving opportunities got away.

I had more than enough high-caliber deal flow to miss a nice deal every now and then. And if the result is fewer losers in my startup portfolio, that helps my overall return.

It was a no-brainer.

But the proliferation of social media data feeds is dramatically changing investor behavior. “Knowing more sooner” is now the go-to strategy.

Better than knowing less later, right?

I have my doubts. And this is why.

A Billion Tweets

There’s no denying that tweets are exploding.

Six years ago, Twitter had generated a billion tweets in all. Now a billion are produced every 48 hours.

Companies like Dataminr are taking advantage.

Dataminr slaps algorithms on its Twitter feed – which follows government- and finance-related sectors – and sells the resulting information to financial customers. It’s now blending its algorithms with its customers to offer them customized alerts.

Does it really work? Is this the beginning of a new era of real-time investment data sources that will only get more sophisticated over time? How can you plug in?

My answers…

Does it really work?

Studies say yes. From a 2011 study by Johan Bollen: Sentiment derived from social media correctly predicted which way the Dow would move three days later 87% of the time.

These findings are supported by a more recent study done by Professor Eli Bartov from the New York University Stern School of Business and two other researchers. It found that tweets before earnings announcements could predict earnings surprises as well as market reactions for individual stocks.

Big-data analytics applied to the wisdom of the crowd is legit. No reason why it shouldn’t work.

Is this the beginning of a new era of real-time investment data sources?

Open to debate. But the short answer is yes.

The longer answer is monetizing these data services won’t be easy. For one, news travels faster now than a year ago but much slower than a year from now.

The advance notice that social media alerts implicitly promise will become harder to come by.

For another: More and more of everything is making its way onto the Web. The amount of information on the Web is expected to increase by five times in the next five years.

A Twitter-like stranglehold on breaking news is fast disappearing. I expect the period we’re in now will be short-lived. And be followed by a more open-ended period when everybody has almost instantaneous access to information captured by social media.

What’s more, social media is susceptible to manipulation. Think restaurant reviews, Craigslist reviews and Yelp.

Some of this manipulation is legal and some not.

How about companies helping clients develop positive social media content and feedback cycles by gaming social media?

It’s perfectly legal and, by the way, increasingly complex, involving a cocktail of audience identification, promotion and advertising, site authority and SEO strategy.

How can you plug in?

Dataminr doesn’t sell to individual investors.

Adam and I use Google Trends. It’s great for consumer-facing companies. Not so good for B2B. More details in this previous post by Adam.

Anybody can plug in for free. But Google searches can also be manipulated. So the online input you get has to be confirmed by more traditional forms of research.

A Deeper Problem

Do you really want to invest in a company getting traction in social media?

The biggest disruptors were misunderstood when starting out. Think Uber. Airbnb. Or even Twitter itself.

At their earliest stages, social media was ignoring/dismissing/ridiculing these companies.

So now we’ve come full circle…

Buffett’s definition of risk – not knowing what you’re doing – has been updated to conform to the new social media age.

Not knowing what you’re doing is a misdemeanor compared to the major crime of not knowing what other investors know (and are talking about).

My take: This is nonsense. More than ever, it’s about what you know – and then some.

Any investor worth their salt should be asking themselves, what do they know that other people don’t?

Game theory says – and hedge funds will soon find out – that if everybody is plugged into competing but basically similar social media feeds, there’s ZERO advantage.

Insight is hard-won. As are exceptional investment returns.

As I see it, tomorrow’s risk will come from blindly following the social media crowd.

Social media signals give you a fleeting snapshot of how a company and business model is perceived. But it’s not a game changer.

And it has a dangerous downside, luring you into exciting and popular investments with uncertain staying power.

Invest early and well,

Andrew Gordon
Founder, Early Investing