Since my open response to Mr. Cespedes last week we have started an email dialog and the following is his response. Please share and comment to keep the conversation going.
First, my sincere thanks to Aaron Mandelbaum for his smart and thoughtful response to my article, “Is Social Media Actually Helping Your Company’s Bottom Line,” on HBR.org. We share the same goal: a more rigorous look at how companies spend time, money, and talent on this activity. There are areas where we agree, and other areas where we disagree, concerning the questions about social media and business performance that I raised in my article.
Where We Agree.
Aaron agrees that few companies have measures for their social media investments; that few companies, according to the McKinsey research I cited, even have accountable managers in place for that spending; that much online discourse about products and companies is fake, bought, or otherwise engineered, not “engagement” with potential customers, and he cites other relevant examples to support this; and that the comScore data about the (un)viewability of many display ads is bad news. His point is that many managers are indeed asleep-at-the-switch with social media and that they should get better at this activity.
That’s a lot of agreement, and should keep people busy—cutting and/or reallocating social media budgets—for some time. If Icebreaker can help companies get better, then you’re helping firms avoid or correct a lot of wasted money, time, and effort. The agreement supports a basic point in my article: “people tend to overhype new technologies and misallocate resources, especially marketers.” If you are a P&L manager or Board member or investor, it also suggests some practical priorities: at this point in the evolution of this medium, addition-by-subtraction may be the first route to explore in examining how your firm uses social media.
Where We Disagree.
I argue that “it’s time to expect more from social media and prove it”—i.e., show links between what is spent here and what happens in buying behavior. Aaron says that digital marketers can infer from likes and tweets the impact on sales. Nope. I cited the accumulating evidence about the weak and often non-existent links between social media activity and sales outcomes, including the essentially random-walk correlation between positive/negative social media conversations and sales predictions (versus after-the-fact tracking of reasons for changes in sales). Let’s see the evidence to the contrary.
I cited a Forrester study indicating that posts from top brands on Twitter and Facebook reach just 2% of followers and only 0.07% of those followers interact with those posts. These numbers make old-time direct mail response rates look good. Aaron says that the “stats are correct but we knew that” and social media allows marketers to “target messages . . . based on expressed interest [and] these are the highest clicked ads on the web.” But these are still clicks, not market outcomes. Lots of people watch TV Super Bowl ads, tweet about them, put them on You Tube, and so on. But we all know that the outcomes—short- and long-term—are very mixed. It’s the old argument about “eyeballs” and, like good performance-oriented managers, I’m asking for evidence about behavioral results, not just cost-per-thousand.
Aaron also disagrees with my questioning the current investment logic in many firms: “Are you saying that a website with 850MM users is overhyped?” Yes, I am. In its heyday, the Yellow Pages reached lots of people, but was that a great ad medium? In the article, I noted that, currently in many companies, the basic investment logic for social media is no deeper than a version of “Fifty million [or more] tweets or likes can’t be wrong.” In business, the road to hell is paved with assumptions like that. Also, despite assertions about hyper-targeting, target demographics on TV are at this point often more knowable and precise than on many demographically chaotic social media sites. Where the better bang-for-the-buck resides is an open question, and the evidence does not necessarily support social media.
Aaron also suggests that, in asking for cause-and-effect evidence here, I am somehow ignoring lifetime customer value or relationship-building efforts or measures of lead generation. Nope. Quite the opposite: I’m urging managers to demand that kind of data in thinking-through their now-sizeable (and often misguided) social media investments.
In turn, Aaron asks, “Have you read about how Target predicted a woman was pregnant?” Yes, I have, again and again. An anecdote is not a business case. And let’s look at that anecdote. As I understand it, Target knew the woman was pregnant (did not predict it) before her parents did. Forget the social and privacy implications for a moment. My question is, from a business perspective, so what? If you market products relevant to pregnant women, there are lots of ways to determine who your potential consumers are. You don’t necessarily need social media to find out who is having babies. The business issue is the relative expense and effectiveness of spending money on X medium rather than Y, not whether somebody at Target happened to use a website to do this.
The fact is, much current business talk about social media is factoids that are oblivious to opportunity costs and a verifiable, programmatic use of the medium beyond simply “having a presence” on Facebook or Twitter or whatever. That must change. And, while “predictions are always risky, especially about the future,” here’s my prediction:
I believe better business use of social media will come from the evolving mechanism of programmatic buying and selling of ads. In these electronic exchanges, buyers meet sellers for all varieties of digital media. It’s like a financial trading desk where data are required for the real-time bidding that takes place. As in any market, the resulting price says a lot about the real value of that ad medium to that advertiser. Programmatic buying and selling today accounts for about 20% of digital ad investments, up from practically nothing just 3 years ago. As it increases, answers to the questions in my article will be necessary for brands, their agencies, and the social media firms seeking to persuade those brands and agencies to buy their wares. That will be a good thing because, as I argued in my article, better use of resources spurs productivity and productivity spurs growth.
Again, my thanks to Aaron and Icebreaker for taking the time and for inviting me to respond.
Frank Cespedes teaches at Harvard Business School and is the author most recently of Aligning Strategy and Sales: The Choices, Systems, and Behaviors that Drive Effective Selling (Harvard Business Review Press).