Understanding the Stochastic Success of Social Media Tactics


We put our video on Youtube so it will go viral. – a corporate executive whose shall remain anonymous to avoid embarrassment.

I actually heard the statement above from multiple times from multiple executives, over the last few months. As many of you know, though, putting your video on YouTube does not mean you have a social media strategy. If only it were that easy. What is likely to happen, really, is that your video gets buried along with hundreds of millions of other videos, and is only watched by the 30 people who receive an email from you with a link to the video.

Social media is all the buzz. Everyone wants to go viral. Every startup I talk to has no marketing budget because, in their words, "we are going to generate buzz among bloggers." Yeah. Sure. You and 247,000 other companies all trying to engage those bloggers for free press. Good luck. Large companies aren't much better. They think social media is just a matter of spending money. Their strategy seems to be:

1. Spend lots of money on a high quality video.
2. Put it on some websites.
3. Reap profits from those marketing efforts.

It's even worse than the strategy of the underpants gnomes, who at least admit that they don't know step 2.

If you want to engage social media, there is one key thing you need to understand. Success is stochastic. This post is going to explain what that means, and I hope it can help you change the way you think about social media strategy.

Stochastic Processes

People tend to think in terms of absolutes. Action X leads to consequence Y. Spend $1 million on TV marketing, reach 400,000 people, sell 4,000 new widgets. It's a forumla, and we like formulas because we understand them. They are deterministic, meaning that for a given set of inputs, the outputs are predictable. Stochastic processes, on the other hand, are driven mathematically, but are not deterministic. They are only predictable in aggregate (technically, this is only true for a subset of stochastic processes, but let's keep it simple). For a stochastic process, you may know that action X leads to consequence Y only 70% of the time. The trouble is, you never know for any given instance if it's one of those 70%. If you run a stochastic process 1000 times, you can predict what you will get. If you run it once or twice, you may hit a homerun, or may get a total flop.

Stochastic thinking is extremely valuable, and it is something you should learn. It can help you better understand financial markets, human behavior, and many natural processes. But it has one big problem…

You can make the right decision and still not get your desired outcome.

Since stochastic processes are non-deterministic, you can never predict exactly what will happen when you give a stochastic system a certain set of inputs.

Application to Social Media

Lucidchart: The Best Value Flowchart Software for Small Businesses in 2016

So that was all boring jibberish, now what's the real point? How can you use this when you think about your social media strategy? Well, the main insight you need to take from this is that you need to do as much as possible because you cannot predict success. Think of it like a probability machine – if you put your video on YouTube, there is a 0.4% chance (I made that number up) that your video becomes popular. Now expand that across all your social media marketing options.

The result, unfortunately for corporate marketing departments, is that you can't really hold your social media team accountable in the short-term, because their results only matter across the long-term. Keeping that in mind, here are my suggestions for marketing departments that want to engage social media.

1. Play – Experiment with new websites, new communities, etc. Encourage people to Twitter or Stumble or blog or whatever it is you want to learn more about. You can't understand these tools and communities unless you use them.

2. Allot money according to risk – Don't go out and spend $100,000 on a super duper Facebook app. Build a scaled down version for $5,000 and test it out, see what you learn, then if you are successful, build on that, if not, kill it off. Put small amounts of money into lots of new and different things. Then pile the rest of your budget into the few big winners.

3. Understand that success is not always repeatable – You can't develop a strategy and stick with it, because the social media landscape is always changing. You have to change too.

4. Understand that failure doesn't mean you made a bad decision – If you spend $20,000 on a few things and you don't get the desired results, it could just be that you fell on the wrong side of the stochastic processes that week. The things that are popular on social media sites have a lot to do with everything else that is going on in the world and on the web. Release your product or project two weeks earlier or two weeks later, and you could get a totally different result. Of course, if you continually fail after many attempts, that does indicate you are doing the wrong things. Which leads me to point five.

5. Look at aggregate results, not individual results – You wouldn't run your business using just a single economic or accounting variable, so don't run your social media marketing strategy that way either. It's just plain lazy.

So there you have it. That's the limit of my knowledge in this area. If you want details, insights, or more than what I've offered, check out these resources for good ideas about using social media tools and engaging online communities.

Paul Gillin
Jeremiah Owyang
Jason Falls
Lee Lefever
Social Media Release
Scott Monty

  • This post is so true! I know clients always want to know what the results will be for their campaigns, but the answer is always “I don’t know”. I can make the best viral campaigns out there, but who knows what will fly and what won’t? The other problem is timing … sometimes a great campaign is slow to pick up, but someone will dig it out a year from now and it’ll be everywhere within a few weeks. The only answer to this problem is to test small, don’t dump the entire budget into one thing, and try often.

  • Very astute analysis, Rob. Thanks!

  • You hit the nail right on the head. The problem is that businesses – and large ones in particular – tend not to reward failure. Your experiment had better work or else. Conventional media is safe and predictable, even if it’s an order of magnitude more expensive. I wish more organizations would embrace the philosophy of “let’s try it,” because trial-and-error are pretty much all the guidance we have to this new world right now.