Turning Facebook Fans into Product Endorsers

V. Kumar, J. Mack Robinson College of Business at Georgia State University

One of the more well-known uses of social business is the marketing and market research function, to try to identify influential consumers on social media sites and listen to what they are saying. Some firms have gone a step further and have tried to figure out how those key people can eventually be leveraged to increase sales for the company. But is there a recommended practice for making this really happen?

V. Kumar, the Richard and Susan Lenny distinguished chair in marketing and director of the Ph.D. Program in Marketing at the J. Mack Robinson College of Business at Georgia State University, says he has come up with just such a process.

Kumar, who also serves as the executive director of the College’s Center for Excellence in Brand and Customer Management, coauthored an article about the process in the Fall 2012 issue of MIT Sloan Management Review. “Increasing the ROI of Social Media Marketing” describes how a seven-step framework for social media marketing helped a small ice cream retailer in India achieve higher sales revenue growth rate, social media ROI and brand awareness.

In an interview with MIT SMR contributing editor Robert Berkman, Kumar goes into more detail on the specific steps and methods that worked best for creating “brand ambassadors” from influential consumers. He explains how companies can locate key influencers, how to link these key people to actual brand growth and sales and other lessons he learned along the way.

Why did you choose the topic of social media marketing to research and write about?

First it is a new area. Second, marketers have a very limited understanding of this area. Third, nobody has measured the ROI of investing in this media for campaigns. So these three factors prompted us to get excited about this kind of topic, and a retailer was willing to work with us because this retailer had a limited marketing budget and wanted to try something out. Hokey Pokey, which is a high-end chain of ice cream shops, was willing to give us access to its cash register for our software to monitor the company’s sales.

In the article, you suggest several steps to effective monitoring, including monitoring brand-related conversations that are happening in the social media platforms. Is monitoring equally applicable to all types of firms and industries?

The tracking of word-of -mouth is broadly applicable to any business because if I say I like Crest toothpaste and I put it as a Facebook “like” and share it with my friends, it can be tracked to determine how many people like it. Then, at some level, you can link the increase in likes to increase in sales.

We see this readily applicable with chain store products. If you are talking about Home Depot and mention that they have a specific product that is great, I can tweet about it. I can send the message, “Get this product from Home Depot, or check out that product which is on sale this week.” If I include a hash tag, and if a customer goes to the cash register and shows the hash tag that came from that tweet, and if at that point of sale the cash register is connected to our computer, we can tie it all together.

As a first step, the sales person has to key in the hash tag when he or she completes the sale. The sale is tied to the hash tag, and our computer makes a reading every night as to which hash tag generated how much sales. Finally, the hash tag is tied back to the individual who spread the word-of-mouth.

Your second step is identifying those individuals to engage in social media conversations. You call the top tier the “influencers,” whose messages are often forwarded and elicit comments. How do you suggest finding these people?

There are many factors. One important way to identify influencers is to make sure they have an interest in your company’s products. For example, you and I could be the significant influencer in talking about a lot of things, and when either of us posts a comment, maybe it gets forwarded to 10,000 people. But if none of these comments are about ice cream, it is pointless for the ice cream retailer to recruit you or me as influencers. That’s where the category relevance comes in terms of the stickiness index when we say what proportion of the tweets that we have pertains to what the retailer cares about. In the case of Hokey Pokey, they were looking for tweets in each ice cream category— milkshakes, sorbet, gelato, ice cream.

We analyze what proportion of the influencers’ tweets belong to this category. Then, in recruiting people, we select those individuals who not only have a higher customer influence effect [CIE] in terms of spreading the message farther, but also a higher share of the stickiness index with respect to this category of ice cream.

We can identify these individuals from our analysis. We approach them and say here is a study, we would like you to participate, and here are the incentives. In the case of ice cream, we contacted about 203 individuals, and brought them to the ice cream parlor in phases. Initially it was about 20 people, and then 30, 40, and gradually the total count increased. And then, we exposed them to the ice cream and the new ice cream creations from this retailer, who started new in India three years ago.

People were excited as everything was so new to them and they made their own creations. You have an ice cream, you have a mix-in, you have a topping, and you end up with a creation. A creation is what we call the outcome of mixing different ingredients. The individuals named their creations, which they marketed to their friends.

Now, do you have to be careful in terms of the incentive you offer? There have been concerns about bloggers that did not disclose payment for blogging positively about companies.

Yes. Fortunately, this is not a monetary incentive. The incentives are intangible: they get to take pictures with celebrities, they get T-shirts and tote bags with the name Hokey Pokey on it. It’s just more of who wins a competition, and the prize becomes very secondary. They all were saying, “I wanted this because I’m more influential among my friends.” So it is the ego that drives the motivation for spreading the word of mouth, rather than any monetary incentive, because that was quite insignificant. So, if I win this week, and I get a $7 T-shirt from Hokey Pokey, nobody can argue that that made me do something.

Is there a way to actually link what you did to an increase product brand growth or sales?

Yes, this is a good question. In statistical terms, you can tease it out if you have a month-by-month list of expenditures in other media also, like radio and television and print. If you have those expenditures, you can bring them into the model, tease the effects out, and then attribute it to just the social media. So that’s the way one would do it, which we have done for many Fortune 500 clients.

But in this case, Hokey Pokey Ice Cream was a new retailer, and it did not have television advertising or print media and radio advertising. The only contribution was social media.

What are one or two lessons that you learned while developing your model that you could pass along to future marketers? What might you do a little differently the next time?

Well, there are two major things. One is that as more and more customers come to buy this product, we now ask them to register with the company so that they willingly give their information to be contacted with promotions or other communication. We seek their permission to track their behavior about their tweets and what they are saying to their friends about the ice cream.

The second aspect is that once they participate, we ask whether they are willing to spread the word and enroll in a program. In other words, as part of the refinement, we are getting more volunteers stepping up and saying, “I want to participate in this because I like this category.”

The reason we give more importance to this is so that we can monitor who is tweeting to whom easily. But if they shut the API interface and then say you cannot monitor, then we need permission of the individuals to enroll and give us permission to monitor their tweets. So we are already thinking ahead to what we call regulated social media marketing.

Are you saying that as we move into the future, it may become more difficult to do the monitoring without express permission?

Yes. So to preempt that, we are asking people to register and voluntarily give us permission. We want them to think, “okay, I like ice cream and these are the benefits I get just from allowing them to track me.” Also, we customize deals for them, not just for ice cream but other special offers, too. In a way, it’s a combination of a social couponing firm like Groupon with social media marketing. A two in one.

There was a study that came out a few weeks ago by the research firm Forrester that found that less than one percent of sales were coming from Facebook and social media. What do you make of that study?

In fact, I wrote an article on this in Forbes a few months ago when GM pulled its ads out of Facebook. It was saying that it was not able to link its advertising messages to any increase in sales. It pulled the impact of a billion dollars’ worth of advertising across all the social media.

It raised a lot of interesting issues and questions. For example, if I go to Facebook at 11 p.m., I don’t want to see an ad for a GM car, I want to just talk to my friends before I unwind. So the relevance of the ads was questionable. Second is the category relevance question. If I’m not even in the market for a car, seeing an ad for a car at that time, I just am indifferent to it. It’s not targeted, that’s a second problem. And third is timing. When I see it, is it relevant?

Even if I say I like an ad and I send it to share it with my friends, they may not be a relevant audience for it either. Everybody has a car. So it is not targeted at all. So there are a lot of problems with advertising in Facebook

But if you engage customers, as we did in our case, it is the customers who are saying that they want to participate in this ice cream social media campaign. So they go and they get their friends to try it out. So it is not an advertisement, it is more a personal selling. And that is a big difference.

It sounds like what you’re saying is that social media marketing, while it may not always be the way to have a direct-click sale, has a wide range of indirect benefits in building the brand.

It has indirect benefits of building the brand, which is certainly intangible, like more word of mouth. In our case, it also increased the awareness of the brand as it increased the positive word of mouth of the brand. So the image went up. All of those things are positive benefits for the brand.