1.
The Social Media ROICookbook:Six Ingredients Top Brands Use to Measurethe Revenue Impact of Social MediaJuly 24, 2012 By Susan Etlinger With Jeremiah Owyang and Andrew Jones Includes input from 66 ecosystem contributors

2.
Executive SummaryToday, customers move constantly between the online and offline worlds, using a range of devices— such as smartphones and tablets — that didn’t exist a few short years ago. Thousands ofapplications and dozens of social media platforms collect and transmit an unprecedented amount ofstructured and unstructured data1, and API changes are a fact of life. The volatility of social data andthe pace of change mean that tried-and-true measurement methods are no longer enough. Socialdata is different. The old rules don’t apply.Although many organizations have established formalized social media programs2, the vast majority— 75% — still lack a holistic measurement strategy3. Web analytics; social media monitoring;social platforms; and tool, application, and ecommerce providers have rushed to fill the gaps, whileanalysts at brands and agencies have borrowed accepted methodologies from adjacent disciplinesto address the unique challenges and pitfalls of social data.As social media matures, new approaches to social media measurement will emerge to provide businesses with agreater level of insight, but the days of certainty (if ever they existed) are behind us. As George E. P. Box4, a notedstatistician, famously said, “Essentially, all models are wrong, but some are useful.”In our research for this report, Altimeter Group identified six primary top-down and bottom-up approaches anddeveloped three case studies that illustrate how organizations measure the impact of social media on revenue. Butwhile these six ingredients are consistent, the emphasis each company places on them depends on the nature oftheir business. There is no “one-size-fits-all” approach. The following pages aim to identify and describe — based onbusiness, product, media, and customer type — the most effective “recipes” for measuring the revenue impact of socialmedia that we have seen adopted to date. Methodology Altimeter Group conducted both qualitative and quantitative analyses, using a combination of online survey, interviews, and briefings, on the ways large organizations measure the revenue impact of social media. Specifically, we conducted: • Interviews with 38 vendors of social media technology, whose products offer some ability to measure the revenue impact of social media • Interviews with 15 brands that are currently measuring the revenue impact of social media programs • Interviews with three agencies • Interviews with four domain experts • Quantitative study of 71 brand and agency-side professionals on measurement experiences and practices Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 1

3.
Table of ContentsBusiness Pain: Organizations Struggle to Quantify Revenue Impact of Social Media ..............................................3Problem: Industry in Transition Leaves Organizations Flying Blind ...................................................................................................3Beyond Revenue: Improving Insight ...................................................................................................................................................7The Six Ingredients to Measuring the Revenue Impact of Social Media ..........................................................................8Top-Down Approaches ................................................................................................................................................................................ 10Bottom-Up Approaches .............................................................................................................................................................................. 11Case Studies ................................................................................................................................................................................................... 13Determining the Right Measurement Mix ...................................................................................................................................... 16The Future of Social Media Measurement .................................................................................................................................... 21Conclusion .................................................................................................................................................................................................... 23Ecosystem Input ........................................................................................................................................................................................ 24About Us ........................................................................................................................................................................................................ 26 Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 2

4.
Business Pain: Organizations Struggle to Quantify Revenue Impact ofSocial MediaProblem: Industry in Transition Leaves Organizations Flying BlindSocial media is no longer optional for business; it’s a fact of life. Today more than 80% of the world’s onlinepopulation uses social media, according to a recent report by comScore5. But social media poses a new series ofinterpretive challenges for organizations, which can make it difficult to assess its impact on the bottom line. AltimeterGroup’s research identified the following primary barriers to measuring the revenue impact of social data:A. Social Media Is Proliferating.Although social media is still very new, Altimeter Group found in the research for our January 2012 report, “A Strategyfor Social Media Proliferation,” that companies average 178 corporate-owned social media accounts. This raisesserious questions: • Do companies know what value they are receiving from these social media properties? • Do they know what resources they are expending to build and support them? • Do they have accurate inventory of these properties?According to the State of Social report by eConsultancy, the answer to each is a resounding “No.” In its report,eConsultancy stated that 41% of more than 1,000 companies and agencies surveyed had “no return of investmentfigure for any of the money they had spent on social channels as of October 2011.”6B. Multiple Challenges Hinder Insight.Fifty-six percent of brands and agencies that Altimeter Group surveyed reported “the inability to tie social media tobusiness outcomes” as the primary challenge to quantifying the revenue impact of social media. But other, moregranular challenges (“lack of analytics expertise and/or resources,” “poor tools,” “unreliable data,” “inconsistentanalytical approaches”) closely followed, suggesting that measurement of social media poses as much of a challengeto organizations and processes as it does to technology. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 3

5.
Figure 1: Brands See Multiple Barriers to Tying Social Media to Revenue“Which of the following have been challenges? (Check up to three)” 60% 50% 40% 30% 20% 10% 56% 39% 38% 35% 30% 0% Inability to tie Lack of Poor tools Inconsistent Unreliable data social media to analytics expertise analytical business outcomes and/or resources approachesBase: 71 respondents responsible for social media measurement in their organizationSource: “The Social Media ROI Cookbook: Six Ingredients Top Brands Use to Measure the Revenue Impact of Social Media,”Altimeter Group (July 24, 2012)In addition to the 35% of survey respondents who noted “inconsistent analytical approaches” as a barrier to insight,several organizations we interviewed highlighted the educational requirements of social data. For example, RobertRoss, Vice President, Interactive & New Media of the American Cancer Society, commented, “We’ve found we haveto help people understand the nature of the data and how to use it and apply it. That’s a delivery challenge.”C. Social Data is New and Different.As outlined in the “Executive Summary,” the complexities of social data make it substantially different from what enterpriseshave seen before. Following are the most commonly cited challenges of measuring the impact of social media: • Exists in online, offline, and across multiple screens, fragmenting data capture. Avinash Kaushik of Google7 has written extensively about the measurement challenges of the world we live in; specifically, the chasm between online and offline and the proliferation of screens through which we interact. This is a significant difference from the early days of the web, before smartphone and tablet devices began to grow in popularity. • Is volatile, making measurement a moving target. Even in the heyday of traditional media, the number of media outlets was relatively stable in comparison to the volatility of today’s social web. For example, from January 1, 2012 to June 5, 2012, Facebook announced the completion of 14 API changes8, while the Pew Research “Twitter Use 2012” report found that “the proportion of online adults who say they use Twitter on a “typical day” has doubled since May 2011 and has quadrupled since late 2010.”9 Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 4

6.
• Organizations don’t “own” social media the way they do their websites or other owned properties. Unlike embedded web analytics on their own website, organizations do not have the ability to tag and track properties on third-party platforms (such as others’ blogs, Facebook pages, or online communities), so they have little direct insight into customer behavior on those properties. • Different apps and platforms give rise to different metrics, making a holistic view challenging. As the old saying goes, “The trouble with standards is that there are so many to choose from.” Facebook metrics differ from Twitter, which differ from Tumblr, communities, and YouTube. To make matters more complicated, new behaviors and features yield new metrics (such as “pins” in Pinterest), and APIs of these third party sites are constantly being updated.Because social media is still immature, it is tempting to think of it in terms of known quantities such as traditional/digital media, marketing, advertising, or ecommerce. But unlike those disciplines, social media is driven from theoutside in, and it happens whether or not we plan for it. As a result, the way we measure social media must adapt tosuit the unique characteristics of the social web. The old rules do not apply.D. Organizations Lack Confidence in Measuring Revenue Impact.Altimeter’s research shows that only 30% of organizations claim to be “very effective” or “extremely effective” atconnecting social media to revenue generation.Figure 2: Few Are Extremely Effective At Connecting Social Media to Revenue“On scale of 1-5, how effective is your organization at connecting social media to revenue generation?” 35% 30% 25% 20% 15% 10% 5% 17% 24% 30% 16% 14% 0% 1 2 3 4 5 (Not at all effective) (Extremely effective)Base: 71 respondents responsible for social media measurement in their organizationSource: “The Social Media ROI Cookbook: Six Ingredients Top Brands Use to Measure the Revenue Impact of Social Media,”Altimeter Group (July 24, 2012)Says Ken Burbary, Chief Digital Officer at Campbell Ewald, “I feel like we’re all data chemists at this point, trying toput a bunch of stuff into our beakers to see if it works.” Although social media has proliferated during the past fewyears, there is a significant gap in the ability to articulate its value. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 5

7.
E. Organizations Are Adapting Slowly.Although social media activity has grown dramatically within the past few years, measurement organizations generallyremain small. Sixty-three percent of brands Altimeter Group surveyed reported that they have one or two peopletasked with social media measurement, while nearly 75% reported fewer than five people in the function.Figure 3: Social Media Measurement Organizations Still Very Small“How many employees are dedicated 0to social media measurement within More than 20your organization?” 6% 13% 11-20 3% 6-10 4% 3-5 11% 63% 1-2Base: 71 respondents responsible for social media measurement in their organizationSource: “The Social Media ROI Cookbook: Six Ingredients Top Brands Use to Measure the Revenue Impact of Social Media,”Altimeter Group (July 24, 2012)Because of the variation in company size, structure, and strategy, these numbers may not necessarily imply thatsocial media measurement is under-resourced. Most organizations Altimeter Group interviewed have taken a cautiousapproach to resource allocation and process improvement as they learn more about the impact of social media onthe business. Says Todd Forsythe, VP Global Marketing, EMC, “We started by saying, ‘Let’s build the organizationalcapabilities, dip our toes in the water.’ Now we’re increasingly becoming more programmatic.”Others brands reported that organizational silos hinder information-sharing, particularly in light of the “inconsistentanalytical approaches” reported in Figure 1. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 6

8.
Beyond Revenue: Improving InsightOrganizations understand that revenue generation is only one benefit of social media and that its main business valueis to deepen relationships with customers and community. While only one of the brands interviewed for this reportadmitted to embarking on a true ROI analysis for social media (analyzing the investment in, as well as return on, socialmedia), all reported that they are making a concerted effort to measure the revenue impact of social media, albeit invarious ways. Survey respondents reported overwhelmingly that the primary business impact of social media was notrevenue generation, but “insight that helped us meet customer experience goals.”The next most-reported benefit was decision-making; 51% of respondents stated that social media measurement“enabled us to make better informed decisions based on social data.” Ali Ardalan, Media and Analytics Strategist atIntel, believes that social media has become a critical input to business decisions and to business cases. Ardalansays, “Why do you do an ROI analysis? To justify why you should do this project vs. another. Why you need morefunding. You need to know the result; are you wasting money? Could you have done the same thing with 20% ofthe budget?”Wes Nichols, Co-Founder and CEO, MarketShare, agrees: “What we’re finding is that businesses will only move bigmoney once they understand what’s truly working. To do that they need a comprehensive view of their marketingperformance, not just the tactical channels.”Figure 4: The Primary Business Value of Social Media: Customer Insight“What have been the primary positive impacts of social media measurement within your organization?(Check up to three)” 90% 80% 70% 60% 50% 40% 30% 20% 84% 51% 35% 32% 26% 10% 0% Customer/ Decision-making Investment Financial impacts Organizational community insight developmentBase: 69 respondents responsible for social media measurement in their organizationSource: “The Social Media ROI Cookbook: Six Ingredients Top Brands Use to Measure the Revenue Impact of Social Media,”Altimeter Group (July 24, 2012)Companies that fail to quantify the hard and soft benefits of their social media programs and activities risk flying blindinto a storm. Those that start now to formulate a measurement strategy, as the companies in the following pageshave done, will be best positioned to succeed in this brave new customer-centric world. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 7

9.
The Six Ingredients to Measuring the Revenue Impact of Social MediaWhile we identified no “perfect” solution to measuring the revenue impact of social media, several best practices haveemerged that are bridging the gaps between what is possible today and what may be possible in the future. AltimeterGroup identified six primary ways that organizations currently measure the revenue impact of social media, whichshould be used as a guideline to determine the most effective measurement mix for your business. The following arebrief descriptions; more detailed descriptions follow. Figure 5: Six Ways of Measuring Revenue Impact of Social Media Top-Down “ Anecdote Specific examples where social media was known to influence a sale or sales. 1. Anecdote Correlation Comparing two data sets (for example, number of likes vs. revenue) to determine whether there may be a relationship. Note that most correlations are quite simple, although companies such as MarketShare are working on far more advanced social econometric models. 2. Correlation A ✓ Multivariate Testing B ✓ Comparing one group exposed to social media content with another that was exposed ✓ to different or no content. C 3. Testing Bottom-Up Links and Tagging Links refer to short links, such as bit.ly, goo.gl, or custom links embedded into content. Tags (and cookies) refer to a piece of code that is embedded into links or URLs for the purpose of conversion attribution. 4. Links and Tagging Integrated Integrated refers to apps or Software-as-a-Service (SaaS) offerings with integrated analytics, such as those offered by Buddy Media, Wildfire, or Facebook apps for Timeline. 5. Integrated Direct Commerce Addition of an ecommerce storefront to a social platform such as Facebook; frequently referred to as “fcommerce.” 6. Direct Commerce Source: “The Social Media ROI Cookbook: Six Ingredients Top Brands Use to Measure the Revenue Impact of Social Media,” Altimeter Group (July 24, 2012) Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 8

10.
The chart below reflects both the immaturity of social media measurement and the resourcefulness of brands tryingto understand its impact on their business in a holistic way. Top-down approaches are as widely used as bottom-upapproaches; anecdote and correlation tied at 44% as the most popular ways to measure the revenue impact of socialmedia, while links were a close second with 42%.Figure 6: Brands Blend Top-Down and Bottom-Up Measurement Approaches“In which ways does your organization measure the revenue impact of social media?” 50% 45% 40% 35% 30% 25% 20% 15% 10% 44% 44% 42% 37% 32% 17% 16% 5% 0% Anecdote Correlation Links TaggingÊ&Êcookies Integrated A/BÊorÊMutivariate Direct Testing Ê ÊBase: 71 respondents responsible for social media measurement in their organizationSource: “The Social Media ROI Cookbook: Six Ingredients Top Brands Use to Measure the Revenue Impact of Social Media,”Altimeter Group (July 24, 2012)What is important to take away from this data is that, while top-down approaches provide business context andbottom-up approaches provide granularity, they are most valuable when viewed in context of each other. The casestudies that follow illustrate how three very different organizations orchestrate these methods to provide insights thathelp drive decision-making. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 9

11.
Top-Down Approaches Figure 7: Top-Down Revenue Measurement Approaches Type Definition and Usage Considerations Sample Vendors 1. Anecdote • Examples of relationship • Lightweight and practical You can use anecdotes • between social media for some companies from any and all social and sales; information • Manual media service or product is often shared verbally, • Cannot scale vendors. This may sometimes entered • No ability to share broadly take verbal or written manually into CRM. form, depending on • Seen in large, often B2B the company. See the companies with high SAP case study for an consideration and long example. sales cycles 2. Correlation • Correlation refers to the • Properly applied, Whatever tool you are • ability to compare one provides insight into the using — web analytics, data set with another to relationship between email service provider — identify patterns. It can social strategies/tactics usually imported into an be as simple as chart and business outcomes Excel spreadsheet. Tools overlays comparing • Well accepted by the such as MarketShare likes and sales to highly industry and Compass Labs complex econometric • Requires trained analysts perform more advanced models that take into who can control for correlations. Expect this account external data, variables and interpret to become more common such as gas prices, data to minimize as tools mature. interest rates, jobless irrelevancies and claims, etc. distortion • Manual (therefore not scalable) 3. Testing (A/B, • In statistics, multivariate • Properly applied, You can compare the • Multivariate) testing or multi-variable provides insight into the performance of any social testing is a technique for relationship between content to any other testing hypotheses on social strategies/tactics social content, either complex multi-variable and business outcomes within the tool itself or systems, especially • Well accepted by digital from tool to tool. Note used in testing market marketing that source data may perceptions.10 • Requires trained analysts come from a variety of • Used by digital marketers who can control for tools, such as listening to compare performance variables and interpret and monitoring, social of tactics across data to minimize media management, different populations; irrelevancies and or any other tool with a examples could include distortion social data output. Some news headlines, social • Manual (therefore not organizations download content of any sort scalable) data into an Excel (video, image, text). Also spreadsheet.11 used by multichannel organizations with both online and brick-and- mortar presence. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 10

12.
Bottom-Up Approaches Figure 8: Bottom-Up Revenue Measurement Approaches Type Definition and Usage Considerations Sample Vendors 4. inking and L • Using a piece of code • Industry standard for Vendors below use links, • Tagging (a short link, ROI tag, or conversion attribution tags, and/or cookies to cookie) to identify the • Enables tracking from identify the source of source of a conversion or click to conversion social content. While sale • Facilitates (measurable) revenue attribution is • At the simpler end: short social spreading a shared feature, they links; at the sophisticated • Only work on owned represent a range of end, ROI tags, such as online properties services, from web those available as part of • Don’t (always) work analytics and digital web analytics solutions across screens marketing optimization to • Broadly applicable to any • Don’t account for social analytics and social online property in which macroeconomic or ecommerce. an item or service is sold business factors (price of Adobe Omniture directly to a customer gas, earnings, etc.) Argyle Social (direct B2C) or a piece of • Only works for single- Awareness Inc. social content culminates browser sessions. Links Badgeville in a desired action, may break easily or may Bazaarvoice such as a contest entry, not be passed through Buddy Media (acquired by white paper download, Salesforce.com) application submission, or Campalyst other desired conversion Exact Target action/lead Expion IBM Coremetrics Interpreter Meltwater Buzz Offerpop Power Reviews (acquired by Bazaarvoice) Revinate ShopIgniter Vitrue WebTrends Wildfire Interactive Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 11

13.
Type Definition and Usage Considerations Sample Vendors5. Integrated Analytics integrated into a • • Highly measurable (if the Argyle Social social media app, widget, conversion occurs within Awareness, Inc. SaaS solution, or service the app) Buddy Media installed on a social • Requires development or Exact Target platform purchase Facebook (Apps for Broadly applicable to • • Managing complexity Timeline) any online property but (multiple apps with Lithium requires development different analytics) Momentfeed of an application or • Metrics may be siloed or Moontoast purchase/use of a tool/ inconsistent with other Offerpop service, such as those metrics Proprietary Tools listed under “Sample Shopkick Vendors” Webtrends Wildfire Interactive6. irect D Addition of ecommerce • • Broadly applicable to any 8th Bridge Commerce storefront to a social online property Moontoast platform (typically • Highly measurable Offerpop Facebook, “fcommerce”) • Limited if you don’t track ShopIgniter full engagement Spiceworks Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 12

14.
Case Studies Case Study: SAP Business: Direct and indirect sale, mixed online and offline onservative culture but with support for test-and-learn approaches C Product: Very high consideration, very long sales-cycle product/service Media: Mixed paid, owned, and earned Customer: Business customer The power of anecdote: Social media frees SAP sales team to focus on lower-funnel activities, improving productivity and close rate. Given SAP’s size and highly metrics-driven culture, one would expect that the company would insist only on quantitative metrics to evaluate the success of its social media initiatives. But SAP values both qualitative and quantitative measures. While it does track more granular metrics (such as numbers of white paper downloads or webinar sign-ups) for lead generation purposes, the company also values qualitative measures that contribute to institutional knowledge. Todd Wilms, Senior Director/Evangelist, Communities and Social Media, reports strong support for social media within SAP: “We have been extremely fortunate from a culture perspective,” he says. At SAP, one of the goals for programs groups and SAP-branded channels is to drive leads. Wilms says that social media has changed the point at which many sales professionals begin to engage with prospects, because it helps prospects to “self-qualify” by engaging with others in SAP communities early in their decision process. As a high-consideration, business-to-business product with a long sales cycle (months to years), SAP is able to track many of the factors that contributed to a single sale. The company has a well-developed customer community (which generates its own far more granular and empirical metrics) with which prospective customers can freely engage. Prospects self-report that they were able to answer their questions and educate themselves through a variety of SAP social channels before engaging directly with the SAP sales team. As a result, says Wilms, “Our sales people can now get involved later in the sales process; they don’t have to jump into the pipeline as early, which has enabled them to have deeper conversations with fewer clients. Then, when they do engage, they want someone to walk them through that last step of fine-tuning SAP into their organization.” Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 13

15.
Case Study: EMCBusiness: Direct and indirect sale, mixed online and offline Culture supports test-and-learn approachesProduct: Medium/high consideration Medium/long sales-cycle product/serviceMedia: Mixed paid, owned, and earnedCustomer: Business customerUsing a mix of quantitative and qualitative methods to prove cause and effect at EMCEMC is taking a holistic approach to measurement, but with quantitative measures at the core. Says KeithPaul, Chief Listener at EMC, “We’re trying to figure out the revenue impact of social. We’re building anintegration and media lab to align everything: social, advertising, search, and marketing sciences.” At the mosttactical level, this means working with marketing teams to develop consistent usage of Omniture codes andbit.ly short links. At a strategic level, it means a shared approach to understanding the impact of social mediaon revenue, as well as on other business goals.Paul has already seen benefits to evangelizing tagging throughout the organization. Recently — using AdobeOmniture codes and bit.ly links — he was able to prove that social media was responsible for generating30% of the viewership for simulcast of a major product launch. Says Paul, “Web analytics is becoming morestrategic again.”At the same time, the company has set its sights on a more ambitious goal: building an ecosystem with usecases and tools that can eventually be used in concert with other enterprise data. Says Todd Forsythe, VicePresident, Global Marketing, “What we’re hoping we can do is connect linking and listening to action in astreamlined way and track — on a real-time basis — the impact of social on business.”In the longer term, this will mean more integration of “Big Data,” specifically, advanced correlations of customerbehavioral data, leads, transactions, and service experiences. While EMC is still in the process of makingmeasurement more programmatic and scalable, the end game, say Paul and Forsythe, is for social data tobecome fully integrated into enterprise data and business strategy. Says Forsythe, “We always have taken theapproach of ‘Plant a thousand seeds, and let the flowers bloom.’” Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 14

16.
Case Study: EventbriteBusiness: Mostly online; service Metrics-driven, adaptive cultureProduct: Short sales cycle Low(er) consideration serviceMedia: Paid, earned, ownedCustomer: Mixed B2B/B2C ighly social consumer and business customer HIntegrating social media into the product provides clarity for Eventbrite.Eventbrite is a company in which social media — in this case, Facebook — is deeply integrated into theproduct itself, making it highly measurable.This wasn’t always the case. “When it was first built, Eventbrite was focused on SEO optimization to drivetraffic,” says Vice President of Marketing Tamara Mendelsohn. “Then in about 2008, we saw that Facebookhad popped up as a top referrer of traffic. Two things were happening: first, event organizers were creatingevents and including links back to Eventbrite to buy tickets. Second, attendees, once they discovered theevent, were sharing the link either pre- or post-purchase. Those two behaviors pushed more traffic.”Eventbrite reached out to Facebook, which opened up the event API to them. It then incorporated sharingfeatures into the product itself. Within a few months, Facebook was the top traffic referrer, which it remainstoday. Recently, says Mendelsohn, Facebook’s mobile site (m.Facebook) began to show up in the top 10referrers as well. While the product is integrated with Facebook, Eventbrite still must use some of the simplertechniques, such URL tracking, to measure business outcomes. While purchases may originate in Facebook,the transactions themselves occur on Eventbrite.The online nature of the business, and the integration with Facebook, means that the company is able toproduce clear, quantitative metrics that demonstrate the relationship between social media and revenue. Forexample, Mendelsohn says, “We can see how many shares and clicks occur and how many of the clicksconvert. And that’s how we derive a dollar amount. Our main financial metric is gross ticket sales. Our topsocial metrics are total dollars driven by sharing divided by total shares, and total visits driven by sharingdivided by total shares.”But transactions don’t tell the complete picture. Says Mendelsohn, “The transaction is the most sacred part of thefunnel, but we’re optimizing all parts of the funnel. For example, if you look at total attendee sharing, 60% of sharingoccurs after the purchase. One in 100 people who look at an event page before purchase share it, while 1 in 10share it after purchase. And a post-purchase share drives 20% more ticket sales than a pre-purchase share.”Ultimately, says Mendelsohn, the question is how to increase the number of shares, especially post-purchase.“If 10% of customers are sharing after purchase,” she says, “how do we make that 20%?” While Mendelsohnis philosophically against incenting people to share — for fear of promoting spammy behavior toward friends— there is still a lot of room for optimization and growth as Eventbrite looks for effective and authentic ways tooptimize the relationship between social and sales. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 15

17.
Determining the Right Measurement MixThere is no universal recipe for measuring the revenue impact of social media. While social media measurement isstill nascent and organizations are still experimenting with approaches, we found that the most advanced companiesconsider the following four factors to determine the appropriate measurement mix: • Business: the nature and structure of the business • Product: the nature and type of products or services offered • Media: type of media being used • Customer: the nature and type of customer(s)Figure 9: Decision Matrix – Assessing the Measurement Mix That’s Right for You 1. Identify Business 2. Assess Product 3. Factor In Your 4. Identify Type Type Media MiX Customer Profile • Sales Channel: • Type: • Paid • Demographics: Indirect or Direct Product or Service (taggable) Where and Who They Are • Delivery Medium: • Sales Cycle: • Owned Online Only or Mixed Short or Long (taggable) • Socialgraphics: Online Offline Where and How • Cost/Consideration: • Earned (not They Interact Online • Culture: Conservative Low or High always taggable) or Experimental • Shared • Target Market: (not always taggable) Business (B2B) or Consumer (B2C)Source: “The Social Media ROI Cookbook: Six Ingredients Top Brands Use to Measure the Revenue Impact of Social Media,”Altimeter Group (July 24, 2012)As social media matures, we will see deeper integration and industry benchmarking that will provide more guidancefor organizations of different sizes and types. For now, however, best practice measurement approaches use theabove decision criteria.The following pages provide guidance on how to determine the measurement mix that is most appropriate foryour organization. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 16

18.
First, Identify Your Business Type.The way your company goes to market largely drives what is possible when it comes to revenue measurement. If your business is … The most appropriate primary Best Practices metrics are … Exclusively online • Links and tags measure revenue • Build in measurement at the (e.g., Zappos, Eventbrite) impact at all stages of the beginning of every significant purchase path. campaign or program. • A wareness: views, impressions • Clearly define goals and metrics • C onsideration: engagement at all stages of the path to (re-tweets, likes, shares) purchase. • C onversion (downloads, • Map volume metrics (views, purchases, registration, shares, retweets, likes, fans, etc.) transaction) with the appropriate stage of the • Integrated: Apps and services purchase path to ensure your offer integrated metrics that you metrics have business context. can use to better understand • Review analytics provided with impact of specific content, apps and services you are channels, or campaigns. considering. The analytics should be a critical purchase criterion. • If you are using multiple apps and services, correlate metrics to better understand relationships. Multichannel • Correlations will help you • If you are using QR codes and (mixed online/offline) understand relationships between barcode scanning, correlate online and offline activities. their analytics with other online • Where practical, use mobile apps sources. to bridge the online and offline • If you are using multiple apps experience and collect data on and services as many companies in-store/offline behaviors. do, correlate metrics among similar apps to better understand relationships. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 17

19.
Second, Assess Your Service or Product Type.The key criteria that drive revenue measurement strategy based upon service and product type are sales cycleand whether the product/service is a high- or low- consideration item. The longer the path to purchase, the lessreliable clickstream data becomes, because there is generally so much elapsed time and so many variables betweenawareness (the reveal of a new car, for example) and purchase (going to the dealer and buying it).Google Analytics has a good way of looking at this challenge from a measurement standpoint. It has divided conversionattribution into two types in their “Social Reports” products. They refer to referrals that lead to conversions immediatelyas “Last Interaction Social Conversions,” compared to what they call “Assisted Social Conversions,” in which a “referralfrom a social source doesn’t immediately generate a conversion, but the visitor returns later and converts.”12 If your product/service is … You should … Best Practices A lower-consideration item (event • Use links and tags to measure • Build in measurement at the tickets, pet food) or has a short revenue impact at all stages of beginning of every significant sales cycle the purchase path. campaign or program. • A wareness: views, impressions • Clearly define goals and metrics • C onsideration: engagement based on path to purchase. (re-tweets, likes, shares) • Map volume metrics (views, • C onversion (downloads, shares, retweets, likes, fans, etc.) purchases, registration, with the appropriate stage of the transaction) purchase path to ensure your • Integrated: Apps and services metrics have business context. offer integrated metrics that you • Review analytics provided with can use to better understand apps and services you are cause and effect. considering. The analytics should be a critical purchase criterion. • If you are using multiple apps and services, correlate metrics to better understand relationships. Longer sales cycles and higher- • Experiment with correlation, as • Don’t disregard anecdote as a consideration items (cars, real these items are frequently sold tool to connect the online and estate) through channels. offline worlds, especially if you are • Use linking and tagging for calls able to log customer verbatims in to action such as white paper a CRM system. While this is not downloads, webinar registrations, scalable, it may yield insight when dealer/agent referrals. you aggregate the data. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 18

20.
Third, Factor In Your Media Mix.The type of medium (Paid, Owned, Earned13) also influences your measurement method, as you can only tag onlineproperties that you control. If the media you are using is … You should … Best Practices Paid • Use links and tags to measure • Build in measurement at the revenue impact at all stages of beginning of every significant the online purchase path. campaign or program. • Print can be linked and tagged • Clearly define goals and metrics via bar or QR codes, while based on path to purchase. broadcast media generally • Map volume metrics (views, cannot, and must rely on shares, retweets, likes, fans, etc.) conventional metrics (or use with the appropriate stage of the correlation, as UFC has done). purchase path to ensure your • A/B testing will show you which metrics have business context. media perform better than others. • Start now to build connections between the digital and social analytics teams. Earned • Because earned media cannot • Map volume metrics (views, be tagged, you must use shares, retweets, likes, fans, etc.) correlations and trend analysis to with the appropriate stage of the understand patterns. purchase path to ensure your metrics have business context. Owned • Owned media (as available via • See “Paid,” above. video, blog posts, social posts, microsites) can be linked and tagged, although only across the same browser. • “Jumping” between screens (phone, tablet, desktop/laptop) breaks these connections, which requires the use of other top- down methods of measurement. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 19

21.
Finally, and Most Importantly, Consider Your Customer Profile.Ultimately, the customer or consumer profile will influence what measurement is possible. If your customer is … You should … Best Practices A business customer • Look at forums, blogs, and • Factor in known social media communities to see where behaviors (Socialmetrics), as well customers congregate and what as behavioral and demographic you can test and learn from data, to help determine how best their interests, questions, and to engage. behaviors. • Test lightweight links and tags in forum replies (as consistent with the behavior of the community) to better understand interest and consideration. A consumer • The decision to use direct • Correlate behaviors at different ecommerce (a storefront) stages of the funnel as Eventbrite should be driven by the type has done, so you can better of relationship you have with understand pre-purchase and your customer base. That said, post-purchase sharing and the deeper the integration of conversion. social and commerce, the more granular the analytics. Highly social • Correlating shares with sales • Be mindful of the balance will provide insight into how between authentic and spammy and whether the online social behavior before incenting experience drives sales. customers to share. • Integrating apps and services • Look at referrers in your web on your Facebook brand pages analytics to find new and trending yields deep analytics. The key sources of traffic (Pinterest, is to translate volume metrics to Tumblr, others). insights and actions (i.e., people who do X tend to do Y). • A/B testing social posts against the same content elsewhere (Facebook, YouTube, your website) will provide insight into how and where people consume your content and where they go (and what they do) from there.Finally, remember that no one set of measures can tell you everything; they all come with trade-offs. But looking atyour granular bottom-up data in context of broader trend data will deliver a more representative view of the revenueimpact of social media than either can in isolation. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 20

22.
The Future of Social Media MeasurementThe ways organizations used to measure web behavior in the past are no longer suited to today’s realities. Thefragmentation of the customer journey (across screens, apps and, platforms) combined with heterogeneous datastreams make it near-impossible for organizations to trace “digital breadcrumbs” from a transaction to its source.14The challenges, as Avinash Kaushik has articulated them15, are straightforward: connecting the customer from onlineto offline and connecting the customer across multiple screens. To this I would add another challenge: the need tointegrate social, enterprise, and external data for a holistic view of behavior and trends. Following are some of thetrends we should expect to see emerge in the next few years.Mobile technologies will bridge the online and offline worlds.Mobile Smartphone and Tablet Apps. While still relatively nascent, smartphone applications such as Shopkick andCheckpoints, which use mobile and geolocation technology to deliver insight on in-store behavior, illustrate thepossibilities of mobile technologies to connect the dots between online and offline interactions.Scanning and Visual Search. Whether it is QR codes, Microsoft TAG, bar codes, or eventually Google Goggles,technologies will continue to emerge that add metadata to offline experiences and enable organizations to view thecustomer experience more holistically. While we are now in a period of feverish development, eventually we will startto see standards emerge in this area. In the meantime, however, the challenge for business will be to select vendorsthat provide the most reliable and stable features and integrations and manage the new data streams these devicescreate without creating undue complexity.Emerging technologies will connect the customer across multiple screens.Matching and Authentication. One of the greatest challenges for brands is the proliferation of browsers and screens,from tablet devices to smartphones, laptops, desktops, and television. In fact, a recent study conducted by Boston’sInnerscope Research found that “digital natives” switch screen 27 times per hour.16Many web analytics vendors aim to connect the social customer using attribution algorithms, while FacebookConnect, Klout, and apps like Empire Avenue attempt to encourage users to self-authenticate so that brandscan “know” them across social channels and platforms. Identity brokers, such as such as Gigya and Janrain andtechnology startups such as Fliptop are uniquely focused on this issue, offering solutions that aim to authenticate andconnect customer identity across the social web.The ability to authenticate identity across the social web also offers an intriguing promise — the ability to connecttransactions to social and interest graphs to better understand social influence on the path to purchase.Data will be the true predictor of influence.As authentication and attribution capabilities improve, the question of who influences whom will become muchclearer. Actual influence (seeing which customers, and customer types, tend to influence other customers to buy)will overtake potential influence (algorithms that predict influence based on fans, followers, and frequency of socialactivity) as a critical business metric. As a result, influence vendors will need to integrate with other transactional datasources to validate their predictions. The winners will be those who use machine learning based on actual businessoutcomes to refine their algorithms. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 21

23.
Companies will integrate social, enterprise, and external data for a holistic view.Social media should not be a data silo; it must be tied to business strategy. The challenge is to normalize social datato the extent possible and integrate it with other enterprise data (such as business intelligence, CRM, and marketresearch) and external sources (such as industry and economic data) to build sophisticated econometric models thatcan be used for modeling, scenario planning, and decision support.We are already seeing movement in this area via industry consolidation (Salesforce’s acquisitions of Radian6 andBuddy Media), Oracle’s acquisition of Collective Intellect, Involver, and Vitrue, and SAP’s partnership with NetBase,although it will take some time to realize the full benefits of these acquisitions.Ultimately, social data, in context of other enterprise data, will become a standard input into business decision-making. The winning technology vendors will be those who seek to solve the whole business problem, rather thanlooking at social as a silo. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 22

24.
ConclusionSocial media for business is still immature, and the mechanisms by which we understand its impact are still evolving.Even if it becomes possible to “match” the majority of people using the social web across platforms, there will alwaysbe those who, for personal, cultural, age-related, security, or political reasons, cannot or will not be identifiable.For that reason, most organizations should expect to use a combination of rigorous top-down and bottom-upmeasurement methods for the foreseeable future, and — to solve the ROI puzzle — will need to start quantifying theirinvestments in social, as well as their returns from it.Whatever ingredients you choose for your measurement mix, the important point is that the organizations thathave been most successful at understanding the financial impact of their social media programs share severalcharacteristics: They are customer-centric, value experimentation, accept that social media is in its infancy and, mostimportantly, have the courage to learn from — and the generosity to share — their experiences. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 23

25.
Ecosystem InputThis report includes input from practitioners, vendors, and market influencers who were interviewed by Altimeter Group during the course of thisresearch. Input into this document does not represent a complete endorsement of the report by the individuals or companies listed below.Brands (19)American Cancer Society, Hilary Noon, VP Customer Insight and ExperienceAmerican Cancer Society, Robert Ross, Vice President, Interactive New MediaCisco Systems, Petra Neiger, Senior Manager, Digital and Social MediaeBay, Sudha Jamthe, Social Media StrategistEMC, Keith Paul, Chief ListenerEMC, Maria Mariotti, Sr. Marketing Programs ManagerEMC, Todd Forsythe, VP Global MarketingEventbrite, Tamara Mendelsohn, VP MarketingHallmark, Camille Lauer, Social Media Insights ManagerIBM, Ranjun Chauhan, Digital and Social Intelligence StrategyIntel, Ali Ardalan, Media and Analytics StrategistSAP, Todd Wilms, Senior Director and Evangelist of Communities and Social MediaStarbucks, Alex Wheeler, VP Digital MarketingStarbucks, Ryan Turner, Director of Global Social MediaThe Coca-Cola Company, Vincenzo Piscopo, Global Director Knowledge and InsightsThomson Reuters, Jaime Punishill, Global Head, WM Digital Distribution and Content StrategyTicketmaster, Kip Levin, EVP, EcommerceUltimate Fighting Championship, Kristin Adams, Social Media ManagerWhirlpool, Stacy Lukasavitz, Social Data AnalystVendors (38)33Across MicrostrategyAdobe MomentfeedArgyle Social MoontoastAttensity NetBaseAwareness Inc. Networked InsightsBadgeville OfferpopBazaarvoice Power Reviews (acquired by Bazaarvoice)Buddy Media (acquired by Salesforce.com) RevinateCampalyst ShopIgniterCompass Labs ShopkickConverseon Simply MeasuredCrimson Hexagon SolariatExact Target SpiceworksExpion Topsy LabsFliptop VinTankGoogle Visible TechnologiesLithium Vitrue (acquired by Oracle)MarketShare WebtrendsMeltwater Buzz Wildfire InteractiveAgencies (5)Campbell Ewald, Ken Burbary, Chief Digital Officer Edelman Digital, Aniz Ruda, Team Lead, Measurement AnalyticsEdelman Digital, Dave Fleet, VP WCG, Chuck Hemann, Director of AnalyticsEdelman Digital, David Armano, EVPDomain Experts (4)Matt Hixson, Co-Founder and CEO, Tellagence Jim Sterne, Founder, eMetrics Marketing Optimization Summit;Marshall Kirkpatrick, CEO, Plexus Engine Digital Analytics AssociationNitin Mayande, Chief Scientist, TellagenceAcknowledgementsWith thanks for support from: Jon Cifuentes, Asha Hossain, Charlene Li, Rebecca Lieb, Chris Silva, Brian Solis, Jaimy Szymanski, Christine Tran,Alec Wagner, Alan Webber, and Susan Wu. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 24

26.
End Notes1 Social networking is the most popular online activity worldwide accounting for nearly 1 in every 5 minutes spent online in October 2011, and reaches 82% of the world’s Internet population, representing 1.2 billion users around the globe. Source: comScore: http://www.comscore.com/ Press_Events/Presentations_Whitepapers/2011/it_is_a_social_world_top_10_need-to-knows_about_social_networking.2 eConsultancy State of Social Report: http://econsultancy.com/us/reports/state-of-social.3 h ttp://www.kaushik.net/avinash/multi-channel-attribution-definitions-models.4 Source: Facebook Developer Blog, https://developers.facebook.com/roadmap/completed-changes.5 Source: Pew Research Center, “Twitter Use 2012,”1 http://www.pewinternet.org/Reports/2012/Twitter-Use-2012.aspx. 6 In statistics, multivariate testing or multi-variable testing is a technique for testing hypotheses on complex multi-variable systems, especially used in testing market perceptions. Source: http://en.wikipedia.org/wiki/Multivariate_testing#cite_note-0.7 For a good overview, see ClickZ, “A/B Testing for the Mathematically Disinclined,” http://www.clickz.com/clickz/column/1704390/a-b-testing- mathematically-disinclined.8 Source: “Capturing The Value of Social Media Using Google Analytics,” http://analytics.blogspot.com/2012/03/capturing-value-of-social-media- using.html.9 See Jeremiah Owyang and Rebecca Lieb, “The Converged Media Imperative: How Brands Will Combine Paid, Owned and Earned Media” for more information. http://www.altimetergroup.com/research/reports/how-brands-must-combine-paid-owned-and-earned-media.10 Altimeter Group explores these themes — The Dynamic Customer Journey, The Adaptive Organization and The Sentient World — in our research. For more about these research themes, listen to our webinars on the topic here: http://www.altimetergroup.com/research/research- themes.11 See Avinash Kaushik, “Multi-Channel Attribution: Definitions, Models and a Reality Check”: http://www.kaushik.net/avinash/multi-channel- attribution-definitions-models.12 Source: Time, Inc.: http://www.timeinc.com/pressroom/detail.php?id=releases/time_inc_study_digital_natives.php.Open ResearchThis independent research report was 100% funded by Altimeter Group. This report is published under the principleof Open Research and is intended to advance the industry at no cost. This report is intended for you to read, utilize,and share with others; if you do so, please provide attribution to Altimeter Group.PermissionsThe Creative Commons License is Attribution-Noncommercial-Share Alike 3.0 United States atwww.creativecommons.org/licenses/by-nc-sa/3.0.DisclosuresYour trust is important to us, and as such, we believe in being open and transparent about our financial relationships.With permission, we publish a list of our client base on our website. See our website to learn more:www.altimetergroup.com/disclosure.DisclaimerALTHOUGH THE INFORMATION AND DATA USED IN THIS REPORT HAVE BEEN PRODUCED AND PROCESSED FROM SOURCES BELIEVEDTO BE RELIABLE, NO WARRANTY EXPRESSED OR IMPLIED IS MADE REGARDING THE COMPLETENESS, ACCURACY, ADEQUACY, ORUSE OF THE INFORMATION. THE AUTHORS AND CONTRIBUTORS OF THE INFORMATION AND DATA SHALL HAVE NO LIABILITY FORERRORS OR OMISSIONS CONTAINED HEREIN OR FOR INTERPRETATIONS THEREOF. REFERENCE HEREIN TO ANY SPECIFIC PRODUCT ORVENDOR BY TRADE NAME, TRADEMARK, OR OTHERWISE DOES NOT CONSTITUTE OR IMPLY ITS ENDORSEMENT, RECOMMENDATION,OR FAVORING BY THE AUTHORS OR CONTRIBUTORS AND SHALL NOT BE USED FOR ADVERTISING OR PRODUCT ENDORSEMENTPURPOSES. THE OPINIONS EXPRESSED HEREIN ARE SUBJECT TO CHANGE WITHOUT NOTICE. Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 25

27.
About Us Susan Etlinger, Industry Analyst Susan Etlinger (@setlinger) is an Analyst with Altimeter Group, where she focuses on social media analytics and strategy. Previously, Susan was a Senior Vice President at Horn Group, where she pioneered the agency’s social strategy offering. Susan is a published translator and has a bachelor’s degree in rhetoric from the University of California at Berkeley. Jeremiah Owyang, Altimeter Partner, Digital Strategy Analyst Jeremiah Owyang (@jowyang) is a Partner with Altimeter Group, where he focuses on social business and disruptive technologies for customer strategies. Previously, Jeremiah was a Senior Analyst at Forrester Research, Director of Corporate Media Strategy at PodTech Network, and Manager of Global Web Marketing at Hitachi Data Systems. He writes the Web Strategy blog (www.web-strategist.com). Andrew Jones, Senior Researcher Andrew Jones (@andrewjns) is a Senior Researcher at Altimeter Group, where he focuses on social business strategy, including how brands can manage social media proliferation. He also follows broader emerging technology trends such as mobile payments and the implications for governance and third-world development. Andrew previously worked in digital marketing and served in the Peace Corps. Altimeter Group is a research-based advisory firm that helps companies and industries leverage disruption to their advantage. Contact Us Advisory Opportunities Altimeter Group Email: sales@altimetergroup.com 1875 S. Grant Street, Suite 680 San Mateo, CA 94402-2667 info@altimetergroup.com www.altimetergroup.com Attribution-Noncommercial-Share Alike 3.0 United States | © 2012 Altimeter Group | 26