Does Social Media Have A Return On Investment?

Do "likes" and retweets add up to sales? Who knows? And who really cares? We're in the I Love Lucy era of social-media marketing, a golden age of unaccountability.

THIS YEAR, AUDI RAN the first-ever Super Bowl commercial to feature a Twitter hashtag. Did you miss that watershed moment? Don't feel too bad: The hashtag — #ProgressIs, a take on the carmaker's line "Luxury has progressed" — flashed on the screen for just a second, near the end of a surreal and entertaining ad that featured millionaires trying to escape from a minimum-security prison, and a cameo by, who else, sax man and Lite-FM staple Kenny G.

In addition to pushing the hashtag on TV, Audi purchased a Promoted Trend ad from Twitter, and it hired Klout, a startup firm that combs through Twitter and Facebook in search of the most "influential" people online. Klout helped Audi find more than 1,100 people to reach out to about the campaign — 200 of them received an Audi travel mug and flashlight. Klout's Audiphiles tweeted more than 12,000 times about the hashtag, creating a viral chain of Audi-related chatter online. The company then chose the best tweets containing #ProgressIs; the winner, @jetsetbrunette, won a trip to California to test-drive some Audis, and she also got to choose a charity to which Audi donated $25,000.

But what did Audi get out of all these influencers' tweets? Did the Twitter campaign prompt anyone to consider buying an A8, say, or to go into a dealership to test-drive one? Did seeing the #ProgressIs tweets at least inspire an outpouring of positive brand feelings toward Audi?

The company doesn't know. "Today the equation to measure that doesn't exist," says Doug Clark, Audi of America's general manager for social media and customer engagement. Audi has a full-time team monitoring its presence on social-media sites, it's constantly posting new content, and it has even held special events for the most devoted members of the online Audisphere. The best Clark can do to suggest that all this work has paid off is offer a study by Visibli, a social-marketing analytics company, which recently found that Audi has the most "engaged" fans of any entity on Facebook. Audi's more than 3 million obsessives apparently outshine even Justin Bieber's minions in their willingness to click the like button.

Clark concedes that, so far, he doesn't have any numbers to prove that all this engagement has resulted in, you know, selling more cars. Amazingly, the company isn't too interested in finding out, either. For Audi, Facebook and Twitter "are places where we know tech-minded consumers are active, where they're seeking to engage with the brand," Clark says. "But can I say that a fan is more likely to buy an Audi? No."

Audi, like almost every major brand in the world, is jumping onto Twitter and Facebook in a big way. EMarketer estimates that 80% of companies will participate in social-media marketing this year, nearly double the number of just three years ago. All of them are feverishly working to get consumers to "engage" — to "like," to tweet, to comment, to share. And they're spending a tidy sum to do so. According to BIA/Kelsey, a media consulting firm, companies spent about $2.1 billion on social-media advertising in 2010; the number is projected to grow to nearly $8 billion in 2015.

The gold rush has inspired a wave of tech startups, like Klout, that are looking to help firms navigate the tricky social-ad scene. These companies promise to monitor and measure the impact of Facebook and Twitter campaigns, and to find the best ways to boost those efforts. Despite this technology, though, social-media marketing often feels like a throwback to the golden age of TV: At least so far, marketers can't predict or measure the impact of their campaigns with anything near the precision they're used to elsewhere online.

What's more interesting is that brands truly don't seem bothered by this. Being on the leading front of marketing while not having to account for their efforts liberates them. "We're trying different ways to help us better understand the 'value' of a Facebook like," says Brad Shaw, Home Depot's VP for corporate communications and external affairs, echoing several other social-media marketers. "But at this point, revenue is not the intent." Applied to social media, William Goldman's famous line about Hollywood would go something like this: Nobody knows anything, and they don't care. You're forgiven for wondering: #ProgressIs? #Really?

Rank And File Popup-Icon

LATE IN 2007, JOE FERNANDEZ, a young tech-obsessed guy, had to get his jaw wired shut for three months while recovering from surgery. The only way he could communicate with his friends and family, he says, was through Twitter and Facebook. But he found his medically imposed silence to be a revelation, rather than an ordeal. "I could tell people my opinion on anything instantly, and the people who trusted me were acting on what I said," Fernandez says. "It hit me that for the first time, word of mouth was becoming scalable and — even more important — the data about all of these interactions were available."

Fernandez quickly began working on a way to tie all these data into a comprehensive picture of each of our online lives. What he came up with was Klout's signature product, the Klout Score, an integer from 1 to 100 that summarizes every person's influence online. The score is determined by a number of factors — including how influential your followers are and how many people retweet or respond to things you say online. It has become, in some circles, an important measure of influence. People are reputedly putting Klout Scores on their résumés, and a few brands, such as Las Vegas's Palms Hotel, are using Klout to identify potential online VIPs for preferential treatment. Justin Bieber, of course, is the king of Klout: He has a perfect 100. Everyone else is second fiddle. Barack Obama gets an 87, the Dalai Lama gets an 86, and Jay-Z struggles with a mere 67. (For the record, I earn a 64 — good enough for "thought leader" status and probably the only time I'll be this close to Jay-Z in any public ranking.)

Klout has now amassed enough data to measure the influence of 75 million people online, and it can slice and dice these numbers. For instance, Fernandez says Klout can identify the most influential people who talk about sneakers in Seattle, or the most-listened-to tweeters on skin-care products in San Francisco. When companies come to Klout looking to target those influencers, the company can track how their messages echo across the social-media landscape. How many extra tweets did Nike get by focusing on those Seattleites?

But what Fernandez can't track is what happens when people read all those comments or tweets: Does the marketing change anyone's feelings about Nike? "I think we'll get there eventually," he says, musing that over time, brands will give Klout — and other social-data-analysis companies — sales information to correlate with online chatter. Still, the problem won't be easy to solve. For one thing, social-media marketing, unlike search ads, catches most customers when they're far away from making a purchase decision. This makes it menacingly difficult for firms to determine what ultimately led to a purchase. Was it something you saw on Twitter or Facebook three weeks ago, or was it the drive-time radio spot you heard this morning?

Wildfire CEO Victoria Ransom is bringing corporate Facebook fan pages to life via contests and sweepstakes. | Photograph by Robyn Twomey
Wildfire CEO Victoria Ransom is bringing corporate Facebook fan pages to life via contests and sweepstakes. | Photograph by Robyn Twomey

That gets to the second reason that social-media marketing hasn't yet proved itself: So far, advertisers aren't asking for any proof, and that limits the ability of firms like Klout to figure out if what they're doing really works. "My life becomes a million times easier if I can show that if you spend $1 with us, you get $1.10 out," Fernandez says. But for many big brands, the amount of money being dedicated to marketing on Facebook and Twitter is small compared with the rest of their advertising expenses. "For a lot of our clients, what they're spending with us is coming out of their 'experimental' marketing budgets," he says. In other words, they don't feel much pressure to account for their efforts. "No brand is challenging us on this. We challenge ourselves way harder than any brand does."

One current alternative is to embrace less sexy, but more Internet-friendly, direct-response advertising models. "Sweepstakes, contests, and coupons have always been popular, long before the Internet was around," says Victoria Ransom, CEO of Wildfire Interactive, another Silicon Valley advertising startup. Like Klout, the three-year-old firm was founded by accident. Ransom and Alain Chuard came upon the idea while running their previous company, a global adventure-travel firm. They wanted to expand their firm's Facebook presence, "but we realized pretty quickly that we were going to have to give people a reason to become fans of our page," Ransom says. The company had run sweepstakes on its site before, but it found that translating those to Facebook wasn't very easy. "We figured we weren't the only ones facing that challenge," she says. The company created a way for all kinds of businesses to create their own promotional applications on Facebook. "Within a few weeks, we'd received calls from both Kayak and Zappos," Ransom says. "We went, 'Oh, maybe this will be bigger than we thought it would be!' "

Wildfire's twist is making sweepstakes and contests social. They're built as Facebook apps, and they're promoted widely on Twitter. Ransom says that Wildfire can often track the success of its campaigns by integrating with its customers' transaction databases. For instance, the firm recently ran a promotion for Jamba Juice that allowed people to collect a "lucky" coupon from Jamba's Facebook page. You'd only find out the value of the coupon if you took it to a Jamba Juice store, and some of the coupons would pay out cash prizes of up to $10,000. The campaign drove tens of thousands of people to Jamba Juice locations; every time someone used one of Wildfire's coupons to make a purchase, the smoothie chain could credit that customer to the promotion.

Still, Wildfire's campaigns suffer from a problem that's common with social-media marketing: Because they're so new, and because they often depend on catching uncertain viral cascades, their performance is difficult to predict. "If we put a dollar in the Google machine, we know exactly what's going to come out," says David Sobie, VP of business development at HauteLook, a Nordstrom subsidiary that runs a members-only, daily-discount fashion site. Sobie has run many campaigns with Wildfire, and he says, "We're often surprised — things that we didn't think were going to take off have been incredibly successful. And others, where all the metrics suggested that something should have been successful, have turned out not to be."

Klout CEO Joe Fernandez, right, is creating an
Klout CEO Joe Fernandez, right, is creating an "influence graph" for social media and then adding brands to the mix. | Photograph by Robyn Twomey

NOT LONG AGO, I was offered a tour of the customer-service department of the future. It's a bright, gleaming space; costs almost nothing to operate; and boasts the friendliest, most knowledgeable representatives in all of American commerce. Where is this call center? And who runs it?

It's online. And it's run by you. Lithium, a 10-year-old company based in Emeryville, California, builds and hosts online discussion forums for companies to let their customers help themselves, and it's one example where social media already seems to be helping companies pay the bills. For companies like Comcast, which have high-profile rapid-response complaint centers, Lithium's technology has revolutionized customer-service operations, usually an expensive part of the business. CEO Lyle Fong estimates that Lithium's work revamping AT&T's online community resulted in AT&T saving 16% on telephone customer support in January 2011 compared with 2010.

We're a long way from Bieberville, but this being social media, no one wants to talk just about minimizing call volume. At Home Depot, Lithium powers a vibrant discussion site where customers discuss home-improvement projects and the products and instructions to use them. "We can look at how your users interact with each other on your site," Fong says, "and we can tell you, 'Hey, here are your community members who are going to be your most passionate fans, and if you treat them right, they're really going to give back.' " At Sephora, another Lithium client, the discussion site has become a place where some of the company's most feverish fans — women who spend 10 times more than the typical customer — log many hours offering advice to everyone who comes along.

Sephora hasn't calculated all the additional sales that this system has generated, nor the labor it might be saving now that its best customers, rather than employees, are answering people's beauty dilemmas. Bridget Dolan, Sephora's VP of interactive media, says that at some point, the company may decide to do just that. Right now, though, "we aren't saying, 'Does every dollar we spend turn into revenue?' " she admits. "No one here is hounding me for the ROI."

Image: Robyn Twomey

Add New Comment


  • Scott Clark

    I've done plenty of social media work and I have learned a few things. If you think people log on to facebook - youtube - twitter because they can't wait to find out what your brand is doing your nuts. If you give something away or offer a crazy deal for your product as long as they click the like button, you will be "liked". Don't even think that being "liked" means you have brand loyalty. Your competition can be "liked" too. And if the brand your marketing doesn't have any type of brand extension that lends itself to being social, ie. people feeling the need to talk about you then you have no reason or business to doing social. Nothing is more annoying than re-purposed print ad copy showing up on my phone or wall. And unless a company makes a really huge "I" they can expect a very small "R".

    Oh, and those people collecting the data on the ROI, well they can make those numbers sing any tune they want.

  • GetMySkin

    Thank you for the thought-provoking article. I found the Audi Twitter hashtag commercial deserving of greater thought. I might even try a similar campaign.

  • scot ennis

    unmentioned thus far is the damage a brand can do by obtaining likes and then boring their consumers via a lifetime of low value status updates. The root of this issue is lack of planning when setting up the success metrics prior to launch. ROI is not that hard to measure, suitable tracking systems already exist, the real issue is that marketers are not been challenged on why, what and how they want to participate.

  • Sanjay Shetty

    I agree with the comment by David B. Thomas, and I feel the challenge people face is in understanding how to translate this into their business as they haven't clearly specified their business objectives. I've found using F.R.Y. ( is a simple way to begin doing that.

  • Ari Lightman

    We calculate ROI as well as many other metrics to justify social initiatives and campaigns in my class at CMU.  In fact we have done it over and over agin for some very large global companies for both internal and external social based projects.  I am in total agreement with Katie in that it's really hard to do.  We have data scientists, natural language processing experts, strategic IT management types and it's still difficult.  You need to focus less on the tools that spit out some # based on a proprietary algorithm that does not port to any other similar application (like comparing apples and oranges) and think about the strategic intent and collection/assimilation of unstructured social data.  There are ways to ensure you will be able to generate the right kinds of activity to connect with a target market and collect data to validate initial hypothesis (If we do A and B - we will see this particular outcome).  However, like all successful corporate strategies, you need to put the right people, disciplines, tactics and technologies in place at the start of the project. If you're interested in how we do it, feel free to reach out

  • Katelyn Friedson

     This is a such an inaccurate view of how mainstream brands consider social media/social commerce/marketing & ROI. Talk to 1-800-flowers, Banana Republic or the New York Times, and ask if they care about seeing numbers. The answer will be a cold, hard, yes.  They want it all, and why wouldn't they? Without knowing what performs, how would you ever run a sustainable business? They want metrics, data, intelligence, performance, actionable insights. Marketing channels may change, but the fundamentals of marketing, analytics and KPIs have not. What genius decided to run this article to illustrate a 'point' using high-end brands than ran superbowl ads! ouch.

  • Robyn Palmer

    Success, simply, is the fulfillment of goals. It all depends on what you'd like to achieve. I always like to consider these questions when planning for a campaign: (1) What do I want the audience to do and based on their personality, decision making, and preferences (2)  how can I enable them to do so (via a combination of these media and message platforms)? 

    Aside from consumer interaction via social platforms, I think there is great value in the ability to mine social media for product- or experience-specific sentiment (because users give it so freely, and buyers trust customer reviews). Consumer product and service providers would be able to align innovation decision making more closely to user needs and therefore better assign resources.  After all, I can tweet and facebook about a turd all day...

  • Ashley from HootSuite

    Part of the confusion with social media measurement is that social media has a vernacular all it's own -- including ReTweets, "Likes" and Klout scores etc. -- which needs to be translated into a common set of metrics in order to align with the rest of the marketing mix. Using the sales funnel is a great way to do this - as already noted by David B. Thomas below. So where do we begin to stack social media against the rest of our marketing tactics to show how it contributes to the bottom line?Clearly, there is a lot of discussion on the topic, and we want to contribute to the conversation (pardon the plug here... ) HootSuite recently released a White Paper series on social media measurement. We invite you to take a look and let us know what you think; do you agree with the tactics? Would you include any others? We want to establish a common vocabulary for social media measurement, progressing its image from a playful pass time to a reliable tool for business. Take a look:

  • Declan Dunn

    You can find ROI in any media if you plan, track, and look for it. Because so many Social Media "experts" come from a PR or branding background, the kind of direct response tracking many of us are accustomed to are often not included, because those methods have rarely included complete tracking to  purchase or participation.

    On the other side are the direct response curmudgeons, who always point to Google's search ROI and dismiss anything less than that as hype or a joke. Social media goes beyond Twitter and Facebook tactics, or coupons, it's a way of doing business, immersing the customer in an experience of your company. If that experience guides the user to a transaction, it can be tracked.

    For me, it's the merging of the 2 worlds of branding and revenue,what I call Brevenue in my own work. Branding develops awareness, creates familiarity, AND promotes product consideration. Let's face it, the direct response dollars that drive online commerce are a teeny tiny part of overall spending, and this is the first wave of a transition of branding dollars to accountability, leading to a sale. These worlds do not have to be separate, and smart companies as noted in these comments are doing just that...does Zappos sell shoes by offering customer service? Of course...

    Moving beyond the old media paradigm of advertising that separates brand from direct response is silly, it's all a process, and it's all the same customer. Branding leads to revenue, and can be tracked, though not by counting Retweets or Klout integrating the customer into a process, sifting out those interested in taking action, and cementing the brand name in the minds of those not ready.

    We're moving people from the social world of pleasure and connecting to the direct response world of problem/solution, and the name of the game is referrals, not immediate sales on first contact. We track conversion of referrals and find this is far more effective than trying to convert users on first's all about the referral, having your audience invite others to participate, that's what makes it social.

  • John Jackson

    My guess is that 9 out of the 10 people that believe in social media advertising/ROI are those selling this hype to others. The reality is that while social media works for a select few, it's a waste of time for far more...especially in the SMB sector. Mark my words, it's only a matter of time before companies wake up and realize that advertising on sites such as Facebook for example are a waste of time/money. People don't go to FB to be sold anything, they go there to socialize. All of you so-called social media gurus better strike now while everyone is buying into the hype because it's just a matter of time before it all comes crashing down. You can claim there's an ROI in social all you want but if that were really the case, most companies would have already realized what a waste of money it really is. The last statement in this article reeks of the dot-com days of old..."no one here is hounding me for the ROI"...seriously??? If that's not evidence that this is all b.s. I don't know what is. 

  • 40deuce

    There's always a way to measure. It really just depends on what your goal is... or if you even have one. From the article above it seems like Audi doesn't really have a goal in mind, they just want to play around in the social space, but that seems to be working for them. However, most companies are interested in know that their efforts in social media are paying off and there are plenty of ways to do so. If you just look through the comments here you will see a ton of marketers who all have different ideas for tracking the ROI of social media. And these are just the few that chose to respond. I work with people who try to show the ROI of their social media everyday and I see them doing so.
    Yes social media is new and shiny, but don't let the shininess distract you from being a real business with real business goals.

    Sheldon, community manager for Sysomos

  • Austin Edgington

    Well, its seems easier to measure ROI based on the campaign's objective. If the Audi campaign's objective was to drive people into dealerships why didn't they as people to visit a dealership and test drive rather than create a viral chain and pick the best tweet. You can only measure what you ask people to do. Contact me Audi:)

  • Ryan Holota

    Measuring ROI is simple in social media, but like any other advertising medium you need to begin with the end in mind. Simply throwing out a hashtag with no clear purpose is no different than sending a piece of direct mail with no call to action. So called social media 'gurus' have advertisers blinded by saying how revolutionary social is, when it is really just another means of communication. The rules really haven't changed!

  • Jenny Sussin

    Same comment I put forth to Katie above, I'm interested to see everyone saying calculating ROI on social is easy. My question is simple then, how do YOU bring in metrics and calculate and track actual sales leads coming from social?

  • JamesHobbis

    Great article, the basic message take away is that house hold brands, like Audi, are jumping on social media - despite that fact that they haven't figured out how to track its impact on the bottom line. Under the "old media" structure, the impact of any advertising on the bottom line has always been tenuous, hence the famous quote "half my ad spend is wasted...". So buying into social media on the same basis is not really a stretch. The key difference is that the impact is in fact measurable, if you set up your sales funnel with measurable actions. Audi could have driven customers to a Audi Facebook page where they could watch Audi trailers, read Audi marketing messages, enter a dialogue with sales or ask book a test drive. And bingo ROI.

    It is so obvious that the question becomes - why wouldn't you want to do that?

  • Babar Bhatti

    ROI of social media is neither trivial nor impossible. As many readers pointed below, it requires discipline and planning. Its a question of getting educated. I've summarized my recommendations in a white paper which shows how structured frameworks help and how to combine listening and engagement metrics to determine ROI. Available here:


  • Jessica MacRoberts

    Completely agree with David below. You can't figure out an ROI after the fact unless you included how you would track the ROI in the strategic planning. there is always a way to track if you work into your "campaign" instead of firguring out how to do it after the fact. People use contest and cupons all the time through fb that generate sale that you can easily track and put together an ROI for.... So bottom line is if you plan for it and know anything about marketing it is possible, otherwise find a new marketing rep!

  • Jeff Waldman

    100% agree with David B. Thomas' comment below.  Aren't business activities supposed to be linked to the strategic direction of the business?  Isn't social media considered a business activity?  I would say so.  Also, I cringe when I keep seeing people use the term "social media campaigns" as if doing something using social media is some kind of an event, or something that occurs at a specific point in time.  Shouldn't it be something that is fully integrated into all interconnected business functions (e.g. HR, Marketing, Sales, IT, etc...) and be ongoing?

  • Robert Madison

    Three people have commented that David B. Thomas left a good comment, but I'm not able to see it?  Does anybody have a permalink to his comment?