A global bank executive recently described to us a challenge for our times. It turns out that a customer who normally would qualify for the lowest level of service has an impressive 100,000 followers on Twitter. The bank isn’t doing much yet with social media and has no formula for adapting it to particular customers, but the executive still wondered whether the customer’s “influence” might merit special treatment.

It’s the kind of perplexing question many companies face as they formulate their thinking about social media. To understand how businesses are approaching the challenge, we analyzed strategies and practices at more than 1,100 companies across several industries and continents, and conducted in-depth interviews with 70 executives who were leading social media initiatives. Our research revealed four distinct social media strategies, which depend on a company’s tolerance for uncertain outcomes and the level of results sought.

The “predictive practitioner.”

This approach confines usage to a specific area, such as customer service. It works well for businesses seeking to avoid uncertainty and to deliver results that can be measured with established tools.

To increase Clorox’s virtual R&D capabilities, the social media team created Clorox Connects—a website that enables brainstorming with customers and suppliers. A typical query posted there: “We’re working on X product idea. What features would you like to see included?” To encourage participation, Clorox uses incentives borrowed from gaming. For example, people who post answers or add rating comments are awarded points. The site features different levels of difficulty, and contributors who demonstrate expertise can advance to problems requiring greater creativity, knowledge, and involvement. The sharpest contributors gain visibility, making participation rewarding and sticky. One early success came after Clorox posted a question about a specific compound for its salad dressings. Five responses quickly came in. The company decided on a solution within a day and brought the problem solver into the product development process.

The “creative experimenter.”

Companies taking this approach embrace uncertainty, using small-scale tests to find ways to improve discrete functions and practices. They aim to learn by listening to customers and employees on platforms such as Twitter and Facebook. Sometimes they use proprietary technologies to conduct internal tests.

The IT services giant EMC is a creative experimenter. It pays particular attention to how its 40,000 global employees use internal social media to locate needed expertise within the company. In an effort to reduce the use of outside contractors, it created a test platform, called EMC/ONE, that helped employees (many of whom were new because of recent acquisitions) network and connect on projects. “We were very clear that in two months we might unplug this and try a completely different approach,” says Len Devanna, the director of social strategy. “This was the reason we were inside the firewall: To be free to make mistakes and learn our lessons before exposing ourselves to the outside.” Within a year EMC/ONE was delivering substantial benefits. For instance, a division that needed to produce a sales video connected with an in-house production group, saving $10,000 as a result. The company estimates that EMC/ONE has generated more than $40 million in savings overall.

The “social media champion.”

This involves large initiatives designed for predictable results. It may depend on close collaboration across multiple functions and levels and include external parties.

Consider Ford’s 2009 Fiesta Movement campaign, used to prepare for the car’s reintroduction in the U.S. It required joint efforts among marketing, communications, and the C-suite. Ford decided to lend 100 Fiestas for six months to recipients who would use social media to discuss their experiences with the cars in an authentic, direct way. It held an online contest to select candidates, carefully choosing drivers with large social media followings. To further reduce uncertainty, it required them to regularly produce content on themed “missions” (for example, volunteerism) and designed a schedule for postings. Within six months the drivers had posted more than 60,000 items, which garnered millions of clicks, including more than 4.3 million YouTube views. The $5 million campaign created a prelaunch brand awareness rate of 37% among Millennials, generated 50,000 sales leads to new customers, and prompted 35,000 test-drives—a level of results that might be expected from a traditional campaign costing tens of millions of dollars.

The “social media transformer.”

This approach enables large-scale interactions that extend to external stakeholders, allowing companies to use the unexpected to improve the way they do business.

In 2010 Cisco launched Integrated Workforce Experience (IWE), a social business platform designed to facilitate internal and external collaboration and decentralize decision making. It functions much like a Facebook “wall”: A real-time news feed provides updates on employees’ status and activities as well as information about relevant communities, business projects, and customer and partner interactions. One manager likens it to Amazon. “It makes recommendations based on what you are doing, the role you are in, and the choices of other people like you. We are taking that to the enterprise level and basically allowing appropriate information to find you,” he says.

Cisco also makes extensive use of video. It conducts most of its training and meetings virtually, through video streamed to desktops and available via video-on-demand. Like Facebook, the system lets users tag and comment on videos. These technologies have accelerated “time to trust” among Cisco’s stakeholders, quickly establishing collegiality and knowledge sharing among new geographically dispersed teams.

Putting Strategy into Practice

The strategies are temporal, and many organizations will progress from one to another. Companies with clear objectives for using and measuring social technologies in a specific part of the organization should begin as predictive practitioners. They should look for a group (marketing, for instance) that wants to become more social in its business.

Creative experimenters are driven in part by small budgets; labeling a project “experimental” can exempt it from ROI constraints. Both the predictive practitioner and creative experimenter strategies can quickly create significant results and learning and serve as a training ground for larger efforts.

Other companies should use—or migrate toward—a larger-scale strategy if they want significant results. A social champion strategy can help companies identify and enlist enthusiasts to expand initiatives inside and outside the organization. As Ford’s Fiesta Movement showed, carefully engaging those who have a sizable influence in social networks can reduce risk.

With all else being equal, the social transformer strategy can have the largest impact on an enterprise, affecting everything from R&D and operations to channel partners and customers. However, moving from a champion to a transformer strategy requires major, companywide changes to such things as incentive systems, business processes, resource management, and leadership styles. The social transformers we’ve seen often have broader social business objectives and view social technologies as a key enabler of—but not the final answer to—those objectives.

It’s worth remembering that despite their ubiquity, Twitter and Facebook are only five and seven years old, respectively. Who knows what new technologies lie ahead? Understanding how company strategies are evolving to use existing social media not only will be of use today but also should guide managers as they adapt to platforms developed in the years to come.

A version of this article appeared in the?July–August 2011?issue of?Harvard Business Review.

H. James Wilson?is global managing director of information technology and business research at Accenture Research. Follow him on Twitter?@hjameswilson. He is a coauthor, with Paul Daugherty, of?Human + Machine: Reimagining Work in the Age of AI?(Harvard Business Review Press).

PJ Guinan?and Salvatore Parise?are associate professors of information systems at Babson College.

Bruce D. Weinberg?is chair of the marketing department at the Isenberg School of Management, University of Massachusetts, Amherst.