Step 5: Measure the Community’s Involvement (Who, What, Where, When, Why, and How)
Do customer communities pay off? Throughout this section, I’ve argued that marketers must aggregate communities to reach and influence groups of people who share similar interests, concerns, or behaviors (or all three). This is not so different from the way companies currently define and segment their markets, except that they usually use demographics, which is a very crude tool for identifying prospective customers.
Now the social web can simplify the targeting task and make it more accurate because people identify themselves (by behavior and interest) according to the communities they join. But do communities really pay off? Research suggests that the answer is yes.
René Algesheimer, an assistant professor of marketing at the University of Zurich, and Paul M. Dholakmia, an associate professor of management at Rice University’s Jesse H. Jones Graduate School of Management, designed a year-long field experiment to find out. Collaborating with eBay Germany’s managers, they identified 140,120 active eBay users, people who had bought or sold something on the site within the past three months but had not participated in eBay’s online communities before.
Algesheimer and Dholakmia randomly selected 79,242 customers and invited them via e-mail to participate in eBay’s customer communities at the beginning of May 2005, offering them prizes such as iPods as an incentive to do so. The other 60,878 customers, who were not invited to join, served as the control group.
So what happened? “Within three months, 3,299 of the invitees became active community participants, posting messages, joining in discussions, and helping other members,” the researchers report. “We call these customers the ‘community enthusiasts.’
An additional 11,242 users became ‘lurkers,’ reading others’ posts without actively participating in the communities themselves.” But what about their behavior? The researchers say they were astonished at the differences between the two groups: “Lurkers and community enthusiasts bid twice as often as member of the control group, won up to 25 percent more auctions, and paid final prices that were as much as 24 percent higher, and spent up to 54 percent more money (in total).”
Enthusiasts listed more items and their monthly sales were as much as six times higher than the control group’s sales. Just as impressive, nearly ten times as many lurkers (56 percent) and enthusiasts (54 percent) started selling on eBay after they joined the customer communities.
For eBay itself, the increased activity of enthusiasts and lurkers “generated approximately 56 percent more in sales during the year that our experiment ran than in the previous year,” Algesheimer and Dholakmia observed. “With a take rate (the fraction of sales that eBay earns as revenue) of 10.3 percent and a gross margin of 82 percent, eBay earned several million dollars in profit from the increased trading behavior of the community participants in the experiment.
Our results show that customer communities pay off handsomely for eBay and suggest that any online company will benefit from nurturing its communities.”
On the Path to Payoff
Long before you start to measure the payoff—in fact, before you actually implement any community initiative—you should clarify your objectives for marketing to the social web.
You might aim for marketing objectives such as:
- Improving customer relationships
- Improving channel relationships
- Building market share
- Building brand awareness
- Inducing product trial
In addition, your financial objectives might include:
- Boosting sales revenue from specific goods and services
- Improving marketing return on investment
And you can certainly use marketing to the social web for societal objectives such as:
- Building awareness of and involvement in charitable or civic activities
- Increasing awareness of specific issues (energy conservation, environmental protection, and the like)
Before you jump on to the social web, be clear about what you would like to accomplish . . . and what measure(s) you’ll use to determine your progress toward reaching your goal(s).
In fact, it’s important to know exactly what you want to measure before you build whatever it is you’re going to build. Before you write a single line of code, before you approve any creative, you need to plan what you want to measure, how you’re going to do the measuring, and how often.
Remember, however, that your objectives and measures must take into account the users’ experience, which is what communities are all about. For example, say your goal is lead generation.
Typically, management may say, “We need to measure conversion at each point in the user path.” Applying this to a social web situation, your management would want to know that user number one entered the site. I’ll call him Larry (although, of course, you won’t know names at this point).
They also want to know that Larry moved through the low barrier registration— that he gave a user name and picked a password . . . downloaded the case study in the site’s low barrier area . . . submitted the additional information needed to gain admittance to the next tier . . . that he went to the next tier and that he completed a form giving the company his phone number or e-mail or more. All that is very possible to do, and it suggests that Larry will be a very dedicated browser, but you have to think through the value proposition.
Is Larry really going to want to give the name of his firstborn child for the honor of providing the company with his e-mail address? This is a common blunder that companies make all the time. I know of one company site where you almost literally have to give your mother’s maiden name just to download a case study that demonstrates the firm’s capabilities.
Sure, the firm is able to measure visitor movement every step along the way, but by doing so it sets up hurdles that defeat another goal: to have as many visitors as possible downloading the company’s material. Users will give you information to the extent they think they are getting something of value in return. This is why you can’t lose sight of the user experience.
Your company’s perception of value is not going to be the same as your users’ perceptions. Senior managers may demand all kinds of conversion numbers before they fund the site, but at end of the quarter they’ll have only a handful of leads if they put up too many barriers. Which do you want? The measurements or the leads? Think about it.
The One-Two Punch
Once again, when you’re considering community measurement issues, don’t forget what you already know. Online measurement challenges are similar to the kinds of challenges you’d face in direct marketing, marketing communications, or advertising.
If, for example, you’re in marketing communications, how do you measure activity now? If you’re an advertising manager, how do you measure ad impact today? I’m very well aware that we’re operating in an era of accountability.
Nowadays, chief financial officers want to know the return on the marketing investment just as they want to know the return on a new building or a new bulldozer. As BusinessWeek wrote: “For years, corporate marketers have walked into budget meetings like neighborhood junkies. They couldn’t always justify how well they spent past handouts or what difference it all made.
They just wanted more money— for flashy TV ads, for big-ticket events, for, you know, getting out the message and building up the brand.”2 So there is a convergence between what management wants and what the social web in theory is able to deliver.
But let me ask another question:
Do companies stop buying television time or print space because they can’t measure, with real precision, the connection between media advertising and sales revenue?
Obviously not. Instead, American companies have settled on measurements, like gross rating points (GRPs) or cost per thousand (CPM), that relate to factors that syndicated research companies, like Nielsen, Arbitron, and MRI, can measure.
Those measures enable outlets to sell their advertising space in a way that feels fair enough for the advertisers who buy them. Yet they’re not really measuring the medium’s effectiveness. When you think about it, GRPs have very little to do with measuring your marketing effort’s effect. Still, that doesn’t mean you should stop buying television time.
Depending on your product or service, your size, and your financial requirements, you probably do need to build some awareness, which you can do using various types of paid advertising.
However, even companies that have little or no television or print advertising can be successful. YouTube.com is a good recent example—no paid awareness-building yet, but it was purchased by Google for $1.65 billion. At the other extreme, the marketing literature is littered with examples of companies that spent millions of advertising dollars and obtained virtually no awareness.
Let me also note that where advertisers can measure the advertising’s effect—in newspaper classified advertising, for example—the Web has had a major impact. I see marketing on the social web as packing a sort of one-two punch, even though there are no cut-and-dried answers to “What should I measure on the Internet and how?” One punch is the digital media marketing piece, in which you build content that is relevant to your marketing goal.
This could be awareness, leads, sales, or some other marketing objective. The second punch is inviting people to participate in, be informed by, or be entertained by that digital media experience. This one-two punch of the social web combines two marketing areas. You might, for example, use it for interactive marketing in arenas where you’re not paying for media placement. (I’m not talking about banner advertising or popups or other intrusive paid ads.)
Or you might use the social web like traditional public relations, in which case you can use Web analytics as you would on the corporate website if you’ve set up the content properly to capture the data.
It’s All about the Data
E-tailers sweat bullets and spend millions on tracking software to make sure they know absolutely everything about the paths users take through their sites so they can optimize the browsing and buying experience. This is a little like (but much cheaper than) following customers through the store, observing the departments they visit, see which displays grab their attention, seeing where they spend the most time, and recording what they buy.
If your company is not an e-tailer, you probably don’t need to spend quite that much on Web analytics or spend so much time designing your site to trace a visitor’s path.
Still, you should look at certain things, depending on your objectives. If your objective is awareness generation or even better, consideration in the purchase cycle, you may want to measure time spent on the site. A number of clothing sites and e-tailers, such as H&M, Levi Strauss Signature, Adidas apparel, Land’s End, and Speedo United States, invite visitors to configure sophisticated virtual models to their body type so they can see how the apparel will look.
Think how this will impact the user experience—and how much more time visitors are likely to spend on a site where they can get a sense of what the clothing will look on them. Another example is car configuration sites that let visitors mess about with vehicle options, color, and more.
A third example would be the consumer electronics hardware sites where visitors can configure their own computer. This last may not be as visceral an experience as trying on clothes or looking at car models, but it is one way to get people to spend time on the site and get involved with the product.
You want to see how well the eventual product is going to meet whatever need brought a visitor to the site in the first place. Once visitors have invested time on your site, hopefully they will either buy from you online (if you’re Dell) or contact a dealer for more information and a price quote (if you’re General Motors).
If you plan to use the site for lead generation, you’ll need ways to measure different things. Where, for example, do visitors come from? Ads on Google or Yahoo? A blog? The link at a partner’s site? Set up your site so you can measure conversions at each step in the process without diminishing the user experience. If we build a microsite as part of our social media marketing program, we are going to do the same as if it were a pure lead generation site.
The ultimate measure, of course, is the number and quality of the leads generated by the site. Or suppose your company wants to reach as many people as possible.
On the Web, you can measure how many unique visitors come to a site during a day, a week, a month; further analysis will help you determine what provokes a flurry of visits. Measuring digital media relations is not so different from traditional public relations in the sense that you’re using measures to evaluate the quality of the coverage you get. Yes, you count the hits (quantitative), but you care more about one outlet versus another outlet (qualitative).
So digital media relations allow a qualitative and slightly quantitative ability to measure the results. In fact, if you set things up properly on the digital media marketing side—and only if you have set things up properly—you can use what are now traditional Web analytics to measure the results: share of voice, share of market, evidence of and quality of community, and cost of market share.
Share of voice is very similar to what you’d measure in traditional public relations. If your organization aspires to leadership in your industry, you have make your voice heard. Be available to respond to editors and reporters from the Wall Street Journal, the New York Times, BusinessWeek, and the trade press who call for comment.
You also initiate discussions about issues You don’t hide in your office. Share of voice is about thought leadership. It goes back to my point that a brand is only as strong as the dialogue it has with customers. If you have more powerful ideas and content on the social web, then you’re going to have a greater share of the dialogue than your competitor.
The more you can explain and promote what has made you successful in a compelling and engaging way, the greater your share of voice. For example a number of technology companies host a forum for customers, chief information officers’s, in the healthcare industry.
Hosting such a meeting where one of the technology company executives asks the questions creates a share of voice on its turf. If you mix professional and customer-created content by inviting a magazine editor or a well-known author who fits that niche or subject matter to you can attract, engage, and keep people coming is an excellent way to obtain share of voice.
Share of market is about marketing to the social web to secure share in priority markets at a faster rate and lower cost per lead than you’d typically get with traditional media. An immediate objective of lead generation should, ultimately, move the company to a goal such as increasing market share profitably. Here’s a hypothetical example.
Suppose a company I’ll call Intergalactic Software (or IS) sees as an opportunity to expand lead generation activities in a specific vertical market that is currently underserved. Intergalactic Software’s marketing department creates a microsite to generate leads and promotes it using digital media relations.
In other words, IS wants to increase share in that particular vertical market using social media marketing. Intergalactic Software may not be able to determine exactly how many points of market share it’s captured because of the microsite, but, then again, it may be able to. If the company has incremental sales that it can trace back to leads generated by this social media marketing, the site will have affected IS’s share of units and dollars in that market.
Evidence of and Quality of Community
This is a somewhat qualitative measure, based on the premise that engaging your constituents, customers, partners, or prospects is a good thing. If you believe that having quality interaction with and between your customers is a good thing for thought leadership, awareness generation, or for innovation is a valuable activity, then the case is made.
If you don’t believe this—or, more likely, you don’t believe the effort is worth the return— then you won’t care about the evidence and quality of the community. In many cases, a key goal of marketing to the social web is to learn, quickly, how your customers want you to improve your products and services. Assume that, in fact, you’ve set a marketing goal of generating new product enhancement ideas.
You could build your community to encourage participants to generate ideas and then you’d count the ideas you actually get. For maximum accountability, measure from idea generation all the way through the product pipeline to see how many and which ideas are actually implemented—and how they fare with customers.
For example, when Kraft Foods was looking into new dieting and healthy products, it consulted the Communispace online community it was sponsoring. Kraft asked online participants what diet food represented to them and how they made choices when they snack.
The corporation learned that customers didn’t feel they needed to deprive themselves or diet—what they really wanted was the ability to control how much they ate. This prompted Kraft to develop 100 Calorie Packs, a line of small, one-person bags of Oreos and Ritz crackers. The results were dramatic: In the first full year of sales, the product line rang up $100 million in revenues.
Tapping your online community for product ideas is a great goal, but there’s more. Let me quote Joe Lichtenberg, vice president, business development at Eluma, a Massachusetts software company with a vested interest in online communities.
Reacting to a BusinessWeek article about online communities, he posted this comment on BusinessWeek Online: “Clearly a great opportunity for companies to get real feedback and advice, but I think what was missed was that these types of communities can be the holy grail for marketers: ‘A person like me’ has become the most trusted source for information about a company or product.
Marketers would do well to create an environment for their customers to interact— and we’re chock full of Web 2.0 tools to make it happen. When marketers think about how to leverage social networking in their marketing initiatives, it should look like more this—a place for interaction and collaboration—and less like a corporate profile on MySpace!”
In other words, once the community becomes involved (especially when it’s a public community), others are likely to say, “Hey, that’s a terrific idea!” You do risk getting a reaction like, “Well, that’s about the most ridiculous idea I’ve ever heard in my life.” But I would argue that just getting a reaction is good—better than being entirely ignored.
You might have noticed that consumer opinion on the Web tends to split widely over books, movies, music, indeed, most products. Look at the Amazon book reviews; they tend be written by people who either absolutely loved the book (perhaps because they’re related to the author) or loathed it. Someone who is simply indifferent won’t bother to say anything. It’s people who have an interest one way or the other who are going to take the time and energy to voice their opinions— that’s evidence of community.
It’s Not All Milk and Cookies
Because the social web is inherently measurable, and because it’s growing so rapidly, companies that offer Web analytics are also growing. According to Jupiter Research, a New York market research firm, Web analytics software and services will zoom from $565 million to $931 million 2009.5 As the demand for online measurement tools has grown, companies have responded.
Google, for example, offers nine Publisher Tools for website owners to use in analyzing a site’s traffic, increasing traffic, and adding interactivity.
The Yahoo Publisher Network offers similar tools. Also, a number of independent companies offer measurement tools, including Ominiture.com, Coremetrics.com, WebTrends.com, and WebSideStory.com. Still, it’s not all milk and cookies—there are real problems in what to measure, how to measure it, and what (if anything) the measurements mean. In an ideal world, you’d like to be able to measure the social interfaces in relation to the number of actual transactions.
For example, suppose you offer a podcast and a discussion with your customers about, say, accounting software. You include a Q&A session with the editor from Accounting Today as a guest speaker. It would be great to know that your podcast produced a record number of requests for information and product downloads.
At this writing, however, such a podcast’s effect would be very difficult to measure. The content I’ve been talking about in this book creates an information circus. When you market to the social web, you have many things going on at once (professionally generated content, user-generated content, interactivity, everything else I’ve mentioned in earlier chapters).
Does this circus of social interaction really have the payoff of pulling people through to try things or buy things? I believe it does, and the research I quoted at the beginning of this chapter strongly suggests it does. But at this moment nobody has conclusively proved that it does.
There is, unfortunately, no metric for measuring the effectiveness of the social interface. High-profile sites such as YouTube and Facebook have been very effective in using social interface to attract vast numbers of people. Will they be effective in encouraging transactions? It’s hard to know right now, although obviously Google’s management believes that YouTube has—or will have—commercial value. So there’s plenty of room for improvement on the measurement side.
I’m a bit amazed that the blogosphere doesn’t have what I would call a media kit. In theory, a blog could study its readers (because it knows who they are), be able to say how many readers it has, and what their characteristics are. At the moment, such measurements are difficult but they may be possible in the near future.
Because the paid media world has decades of experience in measuring and monetizing what works and what doesn’t work in proceeding to transactions, it could help the social web world improve measures for marketers.
Accountability is crucial: Companies want to validate that marketing to the social web is a powerful way to attract and engage a potential customer. Consider the example I used in Chapter 1 of Gillette’s Scruffy Boy trying to convince young men that the unshaven look is not sexy.
The site is a lot of fun, cute, and sexy. But if the CFO at Gillette asks, “What effect has it really had on our sales?” I suspect that making a clear connection between Scruffy Boy and Gillette razor sales will be very difficult. Rather, I believe, Gillette believes that changing customer attitudes will help sales, since Gillette has 70 percent of the razor business.
Do you even need precise measurements? My associate Cinny Little tells me that by simply exploring a website, she can usually sense whether it was effective at achieving the company’s marketing goal, if she can even figure out what their marketing goal was. “If I think the site was ineffective, I will tell you who had too much power sitting around the table—the marketers, the creative people, or the technology folks,” she says. “If there seems to be a good balance between those three parties, the site was probably effective.”
Nevertheless, with social web analytics, companies can learn things they either could not learn before or could not learn as quickly and economically. Consider what’s happening at Compete Inc., an online consumer market research company.
Compete has permission from more than two million people to ask them questions and to observe their behavior online, with the goal of helping marketers improve their products and better satisfy customer needs. CEO Don McLagan tells me that his company can identify trends very quickly because a growing number of these two million people go online to research their options before they open their wallets or sign on the dotted line.
“They look at a car online before they buy it,” he says. “They examine multiple wireless systems before they decide on one. They look for home equity loans from sources other than the place they have a checking account.”
By observing the online behavior of these people, Compete picks up on patterns and gains insights that help marketers make key decisions. Don described how, for an automotive client, Compete evaluated reaction to General Motors’ “Everybody Gets the Employee Discount” promotion. “When something like that breaks you would like to know immediately—if not sooner—whether you need to respond, if you’re a marketing executive at Chrysler or Ford or Toyota or Honda,” he explains. “Should you use a similar promotion or can you respond selectively on particular models? Obviously, you’ll eventually know what happened because the states register car sales and the documents are a public record, but by that time it may be too late for an effective response.”
Two weeks into the first month of the General Motors promotion, says Don, “We saw the behavior of people online and we knew that the entire GM line was not being affected.
We knew Hummers, trucks, and SUVs were being moved, but we also knew sedans were not being moved. Cadillac and Buick were not being moved at all and, to the extent there was movement, it was in Chevrolet with the SUVs and trucks.” Armed with this information, rival auto marketing executives would be able to tell their management, “We don’t need to offer employee discounts across the entire line.”
If Chrysler, for example, had this data, it would know that its popular 300 sedan would sell well without an employee pricing incentive. Do the math: a $400 employee discount multiplied by approximately 20,000 cars equals $8 million that Daimler Chrysler wouldn’t forego in attempting to meet General Motor’s challenge, if it had the right data to make the decision.
Compete’s research suggests that 75 percent of all auto buyers now research their choice online.
Don says these “catalytic consumers” have three distinguishing characteristics:
- Catalytic consumers use technology to make information an everyday resource. They don’t sit on the couch watching television with the family, absorbing uniform marketing messages. Rather, they filter the deluge of messages to find just what they need—and then shut out the rest.
- Catalytic consumers know their value to marketers; it’s hard to miss. These market-savvy consumers are looking for their piece of the action—through rewards programs, downloadable coupons and contests, and other online promotions.
- Catalytic consumers know that they alone have the ultimate power—the power of their own permission. They control when to grant their personal attention, to which marketers, how often, and for what reward. This, in itself, is a very significant development.
If you were an automotive marketing executive with access to Compete’s research, you wouldn’t have to guess about the effect of the General Motors promotion. Says Don, “You know what sort of conversion GM normally gets out of its demand, so you are able to infer what is likely to happen in time to do something about it. You don’t have to wait for the end-of-the-month sales results to make your decision on what you are going to do and be prepared.” What a powerful payoff, indeed.
The button "I like" was definitely one of the secrets of successful Facebook Twitter so - its direct competitor in the field of social networking - have decided to introduce a similar tool in their platform, a button that would take the name of "Star ".