The hot thing in marketing – a field known for following the latest hot thing – is viral marketing. The aim is to develop a message that is to compelling that people spread it, resulting in massive awareness and profits at minimal cost.
There’s a problem here:
Duncan Watts, a mathematical sociologist at Columbia University and Yahoo! Research, has answers – and I’m afraid they’re not too encouraging. “I’ve been using social media to promote my book,” he says, “and it’s just a waste of time – it has almost no impact at all.”
That’s from Tim Harford’s FT column, “Why Social Marketing Doesn’t Work”.
Well, it sometimes does, but we haven’t figures out when or why. Sure, Rebecca Black’s singing of “Friday” went viral and sold a lot of copies, but this wasn’t predicted. She didn’t have a contract. Her mother put up $4,000 for a recording through a vanity recording house (like a vanity press, but for music). So it CAN happen, but doesn’t happen much. http://en.wikipedia.org/wiki/Rebecca_Black
In studying Tweets, Watts and his colleagues find that the source of the tweet matters, but that’s about it in terms of prediction. And 90% of Tweets don’t get forwarded to anybody.
Over at the Messy Matters blog, Sharad Goel (part of the same network of researchers) provides some sample networks and more data based on multiple information sources. http://messymatters.com/2011/07/31/viral/
We find that across the six domains only 1%-6% of adoptions take place more than one degree from a seed node, meaning that the vast majority of adoptions occur either without peer-to-peer influence or within one step of such an independent adopter. Put another way, the cascade structures above account not only for most trees, but also for most adoptions.
In all the examples we study, diffusion seems remarkably un-viral, rarely spreading far from an independent adopter. Our results thus call into question the dominant, epidemic-like models of diffusion, and also the value of viral marketing campaigns. On a positive note, this observation makes life a lot easier. Instead of needing to describe, predict, or trigger a complicated viral process, one can focus on the much easier case of adoptions that spread at most one hop before terminating. It turns out that diffusion is not nearly as messy as you might think.
I’m pleased rather than surprised at this. We are innundated enough by advertising media without it being so easy to have material passed on. Besides, even viruses don’t always spread virally – most fail to spread, just like most YouTube videos.
Even when things do go viral, it’s not clear that helps. Burger King’s subservient chicken? http://www.bk.com/en/us/campaigns/subservient-chicken.html That went viral.
And the creepy guy with the plastic Burger King head? Everybody knows about him, too.
Both were created by the same agency, but the sales results were poor [possibly from multiple factors] and the agency lost the account, proving that awareness isn’t quite everything.
What about “helping” the viral process?
Can we find out what works?
As Harford notes, “Watts would like to see marketing companies running properly controlled experiments to see which messages carry through social networks such as Facebook. But he’s not convinced that they will.”
It’s easier to conduct and evaluate “viral” campaigns if they aren’t truly viral. Here’s an example:
The company incorporated a BzzAgent social marketing program as part of its larger launch plan. While more traditional media was tasked with driving awareness and interest, the BzzAgent campaign was responsible for driving bottom of the funnel impact – trial, consideration and purchase.
The social marketing campaign engaged 8,000 premium coffee drinkers from BzzAgent’s consumer network of influential brand advocates. Participants received samples of two products and pass-along coupons to inform recommendations and spark purchase.
Full story at: http://www.symphonyiri.com/tabid/330/Default.aspx
So, how does this evaluation work:
SymphonyIRI Group conducted a Matched Market Analysis to isolate the campaign impact relative to other media. In collaboration with BzzAgent and the client team, SymphonyIRI selected “test” and control markets. The test markets included BzzCampaign participants and the brand’s other media, while the control markets did not include BzzAgent presence. Comparing test versus control and adjusting for efforts of non-test variables enabled SymphonyIRI to determine the relative sales impact of the program.
The word of mouth campaign generated a reach of over 1.5 million people via in-person and online conversations. And these recommendations were instrumental in influencing consumer purchase decisions.
SymphonyIRI determined that the BzzCampaign was highly effective at driving sales lift. When analyzed over a 13-week period, the test markets experienced a strong 6.3% sales lift for a product with a high base volume of sales.
As part of the analysis, SymphonyIRI calculated the program ROI, factoring in the cost of the BzzAgent program, all outlet adjustments and estimated profit per unit sold. The result was a positive ROI for the company.
Full disclosure: I was not involved in this study, but work in another division of IRI.
Note an important difference here: this isn’t just a “viral” campaign. This is hiring people to spread the word. That’s made easier by social media, but it’s primed by direct payments to people to spread the word, just as you make direct payments to a radio station to broadcast your commercial.