How to measure the health of an online community – guest post by Doug Pruden and Terry G. Vavra

In Customer Experience, Social media by Simon Crompton-Reid

Would you like to contribute an article to Total Customer?  Click here to find out more.

How to measure the health of an online community  analytics How to measure the health of an online community  analytics

Companies that create an online community without also considering ways to gauge how well the community fulfills its members' needs risk wasting money and – potentially worse – damaging their brand.

More and more businesses are looking to form private online communities among their customers to promote dialogue
and/or collect valuable information and insights from them.But in their race to exploit this new tool, organizations are devoting
resources and potentially risking their brands' credibility without disciplined scrutiny of the effectiveness of these communities. We
believe that in much the same way as businesses recognized early in the development of modern advertising that they needed to track the
effectiveness of their advertising spending, so too will today's organizations begin to understand their obligation to carefully and objectively
evaluate the effectiveness of their forays into social networking. As with any new trend, there are specific reasons for customers to
become involved and specific needs they expect from online communities.It appears that in their rush to capitalize on this new avenue
of communication, business organizations may have failed to carefully assess their customers' needs. Instead, they seem consumed with their
own needs and the opportunities these communities may offer them.

Most important needs

Based on our work in social communication, we've conceptualized customers' most important likely needs (Pruden and Vavra 2004).
Accepting these needs, we offer some key determinants of whether a community will succeed or fail based on whether these motivations are
addressed or overlooked. Perceived value of participation. First and foremost, sponsors of private online customer communities must recognize that the
only reason a customer might continue to participate in a community is if they feel they are receiving some form of value in return for
their involvement. We believe that customers are relatively rational in according attention or participation to activities. They choose to
participate in activities that in some way enrich their lives by adding value.

Ease of use.

Customers will not "endure pain" to participate in a community. This means the portal and all aspects of participation
must be as easy to use as possible. Not only must they be easy, but interaction with them – a major
determinant of the customer's experience – should be painless or even enjoyable.


Many of the customers who volunteer to participate in a private online community will be doing so in the hope of learning –
learning more about a product or brand, becoming experienced users of a technology or product category.
Their ultimate goal might even be to become "power users." And so they will look to their participation in the
community to further this learning objective. The community sponsor needs to ask itself, "How much
information have we transferred over to community members to answer this need?"

Part of the inner circle.

There can be something quite exclusive about participating in an online community.
Members can be made to feel like they're part of the action. This is especially true of communities in which members are asked
their opinions or advice on product planning or development. And being involved with specific glamour brands can also serve as a substantial
motivation. Imagine the ego boost of receiving personal e-mails from corporate giants like FedEx or Nordstrom! In apparent recognition
of this, Neiman Marcus has, for some time, branded its communications with its best customers as "The Inner Circle."

Making friends and acquaintances.

Social networks are all about making new friends – often, many of them – in cyberspace. And, in a very
unique way private online communities fuel this need by enforcing some degree of similarity among the members, making the sense of community
even more desirable. An imported automobile manufacturer's community of Gen Xers is a good example.


Most humans look to one another for acceptance and respect. Joining an online community can be perceived as an important way to gain esteem. Being invited by a major manufacturer and then accepted within a community can be a real ego trip. It is critical that sponsors understand and feed this need.

Seem oblivious

Strangely enough, many community sponsors seem oblivious to these motivations (Nedelka 2009), being content with milking their
customer-participants for involvement, ideas and time without any thought of what they are giving in return or how they can enhance
the experience for their participating customers. Exploiting customers' participation is a surefire recipe for early decay of a community. There is
absolutely no reason for customers to continue to participate in a community if their experience is neutral or,
worse yet, unpleasant. Because of this disregard, attrition Рonce the bane of many subscription industries (pagers, cellphones and cable TV) Рis quickly becoming a challenge to the social networking industry. We believe the genesis of this attrition is organizations'blas̩ attitudes toward enriching the
experiences of community members.
Even those sponsors who worry about attrition lack the tools to properly understand the reasons for member defection.
Despite the intense interest in using private online communities, little has been discussed about how to assess their true contribution to an organization's marketing plan (Paine 2007). To answer the need for accountability in the management of online communities and to foster
better management on a real-time basis, we propose the following community engagement index.
This research and planning tool summarizes the current participation of community members; measures community members' attitudinal
connection with the community; predicts future participation levels; and identifies which management tactics are working best.
The measurement tool builds from our writing and experience in relationship management, monitoring of word of mouth and our conceptualizing
in person-to-person advocacy (Pruden and Vavra 2010).
Oversight of private online customer communities provides all sorts of participation-based information. The statistics that can be collected on members' visits and involvement (e.g., frequency of visits, duration of visits, and "quality" of visits – degree of contribution) are readily available yet are often treated as discardable.

Rarely are they properly interpreted for the insight or trends they might yield. But this plentiful information also poses problems. The items provide
a look in the rearview mirror rather than a projection of where a community is heading. They describe reality but fail to provide an understanding
or explanation for the behavior that's being observed.
To overcome these hurdles, the ideal engagement index should complement observable data with attitudes collected directly from community members. These data points, which we call "emotionals," add considerable value to the observables through their ability to help explain movements and trends within the
observables. For example, consider a community in which the number of active members is decreasing.
Observables can only report that attrition is increasing; they can't tell us why. But, if the proper emotional measures are also being collected on
a periodic basis, the sponsor will have considerable information from which to diagnose problems and from which to quickly make remedial
changes in the management or tasking of the community.

Not recognized to have value

Most managers of private online customer communities have available (or already collect) a wealth of behavioral information. However, in most cases
this information is not recognized to have the value it really offers. Instead, this information is considered a by-product by the technical staff who are managing the communities.
Their goal is to improve the functionality of the community; they are not held responsible for anticipating problems or for improving experiential components of their communities.
Typically the following will be available:

Population. The total number of members enrolled/empaneled in the community at the time of formation.
Defection/Defectors. To appropriately derive this measure, active and lost members need to be precisely defined. Community membersfailing to exhibit the described activity (e.g., must have participated in at least two of the previous four events) should be considered lost, inactive or defectors.
Replenishment/Replacement. The cumulative number of new members actively recruited by the sponsor or manager to replace members lost through attrition. This will be reported on a percentage basis.
Attraction. The cumulative number of new community members who appeared through member referral or other word-of-mouth recommendation
through an observation period.
Participation. Implying a measure of activity, participation can be derived by dividing the number of unique logins by the current population.
Involvement. This construct suggests a measure of global productivity. It is usually defined by a count of the total contributions from all members of the community as a whole during a reporting period. Total contributions is composed of: total chats + total poll responses + total posts/submissions +
total ratings + total comments + total votes. Community members may be counted multiple times within this measurement based on participating in
numerous ways.
Interest/Dedication. While involvement is a binary measure (did or didn't happen), interest is a more quantitative measure implying commitment to the community. It is generally defined as the average length of time per login for all members of the community.
Number of sponsored events. A tally needs to be maintained of the total number of sponsored events/ invited activities conducted during the report time period.

Qualitative feelings

What most managers of private online communities lack is more diagnostic information about the qualitative feelings of customer members who are participating in the communities. Our emotional measures deliver that feedback. Linked with the behavioral measures, they offer an explanation for the behaviors and can provide real direction for how situations can be improved upon. The only drawback is that emotional measures need to be consciously collected in conjunction
with other interactions with the community. In this respect they ask more of community members. Sometimes community managers will worry that the additional burden will be too much. However, considering the value of the information, the inconvenience imposed on members is more than justified.

Here are our emotional measures:

Value. Continued participation in any community is going to be strongly influenced by the extent to which its members perceive value accruing to them for participating. Far too many communities are an egotistical exercise by a company or brand that fails to understand the necessity of providing value to the
members as well as to itself.

Intellectual engagement. While perceived value is necessary, the ideal community will stimulate thinking about the sponsor's product,
industry or reputation.
Enjoyment. Value and intellectual stimulation can occur in a relatively sterile environment, so we believe it's also desirable to measure the entertainment value of a community.
Trust. In an era in which many consumers feel their privacy is being compromised by information that's being collected over the Internet, the prudent community sponsor will track community members' comfort with providing information to the community.
Usability/Ease of use.
Navigability and ease of use substantially impact interaction with online communities. It's valuable to know from the member's perspective how easy it is to use and interact with a community Web site/portal.
Advocacy. Some online communities may be created purely to collect information. It would be difficult to conclude that these communities
should have any positive impact on feelings for the sponsor, the sponsor's brand(s) or the sponsor's industry. Yet other communities may hope that positive associations from participation will transfer to the sponsor. The most beneficial form of  positive impact would be for a community
member to talk positively about the sponsor or sponsor's brands to friends or acquaintances. We refer to such behavior as advocacy.

engagement indices online community analytics

Help develop hypotheses

Ideally, reports on the behaviorals and emotionals will be developed monthly or weekly depending on the expected life of the community and the frequency and volume of community events. Trends emerging among the behaviorals and emotionals should help community sponsors develop hypotheses about successes
or failures. Most importantly, by projecting unfortunate trends and understanding their probable causes, the astute community manager will be
able to correct problems and reverse the trends.
Beyond the monthly diagnostics, the measures are combined to form a single index. This index provides a useful measure for community managers with which to easily compare the performance of numerous communities. In doing so, successful management techniques and managerial practices can easily be identified. In an ongoing process the weights of the individual measures (building to the index) will be constantly evaluated using a success criterion as a dependent
variable. Weights are expected to vary by community and by industry.
We believe that the most important use of the index should be in helping to improve the customer experience in maintaining active, healthy participation of customers inthe community.
To demonstrate the insight provided by the community engagement index, Figure 1 presents data illustrating what three typical communities
(Alpha, Beta and Gamma) might look like. First, the communities may be compared using the cumulative 12-period index to get a sense of their comparative health. Community Alpha has a 6.8 cumulative index, Community Beta a 5.5 and Community Gamma a 7.0 cumulative index. From this comparison, it would be concluded that Community Gamma is, overall, the healthiest, based on the total components of the engagement index (the observables and the emotionals).
A second way to use the engagement index is to help determine what may be going wrong in a community. Community Beta has the lowest cumulative engagement index (5.5). Reviewing the behavioral measures across the three communities, it becomes evident that Community Beta's participation rates are the
lowest. Further, its participation rates are declining rather substantially (from 65 percent-35 percent-83 percent in the first three periods down
to 35 percent-51 percent-31 percent in the last three periods). The supervising management team may have some thoughts on how to explain the
decline. But with the emotional measures of the engagement index, there can be a more objective explanation. Looking down the average scores for
the emotionals, the low scores (6.4  and 6.1) for intellectual engagement and enjoyment suggest a possible cause: The community is apparently not providing an engaging, enjoyable experience to its members.
A third use for the engagement index is to track a certain outcome or property of a community. As an example, some clients and community managers believe that hosting a community can have a very real benefit to the sponsoring brand or company through the positive word-of-mouth that's created. Looking at
Community Gamma, the healthiest of the three communities (a cumulative engagement index of 7.0 compared to Alpha's 6.8 and Beta's 5.5), it has one conspicuous weakness: It is far lower in the emotional measure of advocacy. Despite its overall success, Gamma's scores for advocacy have dropped steadily throughout the 12-period life of the community. At a period-12 rating of 3.3, Community Gamma's advocacy likelihood is approximately half
the other two communities (7.7 for Alpha and 6.0 for Beta). While it may be entertaining its members, it's not as successful in generating word-of-
mouth for its sponsoring company or brand as the other two communities.
This is an enlightening discovery for management that wouldn't have been evident given a more traditional examination of just the observables.
Whichever of these three analytical perspectives management employs, the engagement index holds great potential to foster more proactive oversight of private online customer communities through the data it can provide.

Disciplined monitoring program

Private online customer communities have tremendous potential as a part of a company's or brand's marketing program. To be truly successful, and fully support the brand and the customers, a disciplined monitoring program that tracks both observable usage information as well as emotional responses from participating
customers must be part of the process. With the insights gathered through such a monitoring program, communities can be better managed and populated with content and activities that meet the specific needs of community members, thus insuring the long-term success of the community and maximizing the benefits
the sponsor receives.

Doug Pruden and Terry G. Vavra are partners at research firm Customer Experience Partners.

Based in Darien,Conn., Pruden can be reached at 203-655-0090 or
Based in Allendale, N.J., Vavra can be reached at 201-825-2556 or at

Pruden, Douglas R., and Vavra, Terry G.
"Controlling the Grapevine: How to Measure
and Manage Word of Mouth." Marketing
Management, July/August 2004, pp. 25-30.
Nedelka, Jeremy. "Measuring the Loyalty of
the Social Customer." 1to1 Magazine. Winter
Paine, Katie Delahaye. "Three Approaches
to Measuring Customer Engagement." The
Measurement Standard. August 19, 2010.
Pruden, Douglas R., and Vavra, Terry G.,
"You're Losing Control of Your Brand Image".
DestinationCRM. June 1, 2010.

Would you like to contribute an article to Total Customer?  Click here to find out more.