Marketing Pilgrim's "Inbound Marketing" Channel

Marketing Pilgrim's Inbound Channel is sponsored by SDL. SDL Campaigns empowers marketers to create, deliver, monitor and measure dynamic, highly-targeted and personalized digital marketing campaigns across all channels and devices.

Science, Not Guesswork, Will Bring Social Media Out of the Dark Ages


For a full hour of ground-breaking social media science, data and myth busting, be sure to register for HubSpot’s Science of Social Media webinar. HubSpot is the sponsor of Marketing Pilgrim’s Inbound Marketing channel.

It’s easy to get sucked into soft-focus, touchy-feely social media advice. Stuff like “engage in the conversation,” “hug your followers,” and “have a personality.” And it’s often not easy to disagree with myths like these, since they “seem right” and nobody wants to be the guy that says “punch your customers in the face.”

The same thing was happening with medical advice centuries ago. “Common sense” approaches to health care led to things like magical tonics and bleeding. Not only did these things often not work, but much of the time they actually did more harm than good.

But then medicine came out of the dark ages through the scientific method and practitioners and researchers began to use real data to understand what worked and what didn’t. Beyond “common sense.”

Social media needs to come out of the dark ages now too. We need to start using the mountains of real data and the scientific method to conduct real experiments to learn what actually works.

Science isn’t perfect. Neither are statistics. Correlation and causation, tortured data and faulty experiment design are all real dangers. But even questionable science is better than just guessing.

Relying on “years of experience” is just as dangerous as sheer guesswork. Years of experience is basically using a very small data set, with serious selection bias issues to generate anecdotal evidence and then running with it. It’s better than nothing, but it’s nowhere near as good as objective research that uses sample sizes north of tens or hundreds of thousands or even millions of data points.

And fuzzy, nebulous advice like “personality is important” or “be interesting” is totally worthless. It’s like saying “be cool.” It’s far too broad, subjective and weakly defined to actually mean anything. This kind of advice doesn’t give anyone real direction. It just “sounds” right, and so it’s easy for the masses to cheer for.

But we’re in business to make money, not friends, right? I’m not interested in what’s popular, I’m interested in what works. Are you?

Dan Zarrella is the award-winning social media scientist at HubSpot.

  • Tommy Swanson

    Couldn’t agree more, Dan.

    I sometimes feel like we’re in the dot com era, where you just “have to have a website.” Now, “you have to have a facebook page and twitter profile”?

    Why? “Because everyone is doing it.”

    We need to start digging deep and figuring out what works and what’s driving revenue or whatever you’re trying to accomplish.

    When data starts to drive your strategy, that’s when you’ll get significant buy-in from organizations.

    I know you’ve been getting a lot of crap for your research lately, Dan, but standing up for what’s right, even when you’re alone, is what differentiates yourself downstream.

    You’re the man, Dan! :)

    • http://danzarrella.com Dan Zarrella

      Thanks dude. I’m trying, but it’s not always easy.

    • http://netmajicsocialmedia.wordpress.com Brian Johnson

      I think the sentiment that “you have to have a facebook page and twitter profile” is not necessarily incorrect for most companies, though it’s not because “everyone is doing it.” I would say that it is true more because “it’s very easy to do and people are looking for you there.” There is a lot more to social media than that, obviously, but isn’t that reason enough? If a potential client or investor or whoever is searching you out in those places, and its not that hard to be there, then you should probably make sure you are there.

      There are certainly exceptions to everything, but for the most part, you may as well be there.

      That being said, as far as running actually social media campaigns and the strategy behind being actively involved in social media, I totally agree, data and scientific evidence should drive everything. And yes, you cannot just tell someone to “have a personality” as if telling them that is going to instantly lead to a more successful campaign. Social media as well as many parts of marketing need to go from the intangible and unsubstantiated to the concrete and proven.

  • http://www.kilkku.com/blog/ Ville Kilkku

    I agree that we need proper studies and proper data on social media, just like on everything else. This is very important work.

    However, one sentence really rubbed me the wrong way: “But even questionable science is better than just guessing.”

    I do not think this is true.

    The social space is full of very questionable statistics: Either the methodology is completely bonkers, the sample set is horribly biased, or the sample size is so small that the margin of error is somewhere between 10 and 20 percent, especially when trying to segment the data. In worst cases, a single statistic suffers from all three.

    The problem with messed up statistics versus guessing is that messed up statistics assert a pretense of proof, whereas they are no better than guessing. This makes them more harmful. When compared to significant experience, they just lose flat out.

    In short, I would choose years of experience over questionable science any day. But I would never choose years of experience over diligent science.