Facebook Retail Sites Dying, And For Very Good Reasons

When stories hit last week pointing out that several major chains—including JCPenney, Nordstrom and Gap—had killed their Facebook sites due to insufficient new revenue, some started questioning the value of Facebook and social media in general. It was a wonderful example of drawing the dead-wrong conclusion from research.

Those Facebook sites died because they were based on a flawed understanding of what social media is all about. It's not about creating a storefront or a virtual watering hole where customers gather to sing your praises. Those retailers already have that: It's called a Web site. Social media is about listening and only participating when you are either invited or you can be truly helpful.

With social media, return on investment (ROI) has to be viewed from a very long-term perspective. That doesn't play with the marketing groups who are trying to create these Facebook areas.

"Candidly, [waiting for ROI over the long term] is just not an interesting approach for marketers," said Forrester analyst Sucharita Mulpuru. "They want to be able to say that 80 percent of the traffic comes from Facebook. Agencies drive that, because it gives them a reason for being."

Social media delivers one attribute that is absurdly powerful, but it requires tremendous discipline to use: credibility.

Remember those stealth viral marketing campaigns that were popular with bars? A beer company would pay people to go into bars, order their product and say nice things about it to others. Obviously, the whole effort would fail if that plant wore a sign saying he's being paid by that beer company to say these things. A Facebook site for a specific company is kind of like that guy in the bar wearing the beer sign.

The power of social is sitting back and letting your true customers talk about you in the context of wherever they naturally want to go. Any attempt by a retailer to try and influence that conversation risks blowing the whole opportunity.

The ability to gather intelligence is huge, as is the wonderful ability for your people to be there—lurking—to answer a complaint or to help solve a problem. You can get yourself some amazing loyalty, if you're restrained and polite. If you weaken for just a moment and make a sales pitch, and you've destroyed weeks of building good will.

Don't give up on social. But also do not treat it like the Web or anything else.

Even Likes are problematic. The assumption behind Likes and Friends is that these people endorse your brand and that a large number of them will encourage purchases from other people. That only works if the numbers are credible. Campaigns to bribe people—out in the open—to Like or Friend your brand are going to make people dismiss those numbers.

Some of the differences of social have been quantified, but it can be hard. When the media approach is as different as social is, using the old research methodology can be ineffective at delivering meaningful results. Indeed, the very nature of social resists easy calculations.

Take, for example, a company called Klout.Take, for example, a company called Klout. Klout generates ratings for various companies (retailers included) about how effective they are in social media. To calculate this, it looks at retweets and mentions on Twitter; comments, wall posts and Likes from Facebook; comments and Likes from LinkedIn; Tips, to-dos and Dones from Foursquare; and comments, reshares and +1s from Google+.

Every analysis metric has to start somewhere, and this seems like a fine place to begin. But what I would much rather see on that list is phrase and concept tracking. If Macy's tweets something that makes me think I should re-evaluate how long it will be until we replace our bed, I'm likely to ask friends or associates about it. But I'm not likely to retweet the Macy's message nor even acknowledge that my question is based on a marketing blast.

I'd rather see how many conversations about bed replacement happened after Macy's message went out. Specifically, I'd like to see a before-and-after stat, as in "For 90 days before the tweet, we saw an average of from 9 to 14 Tweets on the day on that subject. Starting 10 minutes after that tweet, we tracked 8,900 such tweets." That to me is meaningful, not mere retweets or Likes. Consumers don't want to feel used, and redistributing marketing messages does that.

The Klout problem is like that very old joke about the guy who notices a stranger standing by a streetlight, carefully examining the street below. He asks the stranger what the problem is and the stranger says, "I lost my car keys." The guy replies, "Where exactly were you standing when you lost them?" The stranger points about 400 yards away.

"If you lost them over there, why are you looking here?" The stranger replies: "The light is better over here." Just because the analytical light is good at counting up retweets and Likes, that doesn't mean those results will help you find your revenue boost or your car keys.