Getting the Word Out in the Last Green Valley: Integrating Digital Video, Direct Mail, and Web-Based Information for Specific Target Audiences

Introduction and Background

The Green Valley Institute is a formal educational partnership program among the University of Connecticut Cooperative Extension System, the University of Massachusetts Cooperative Extension System, and the Quinebaug-Shetucket National Heritage Corridor. The Heritage Corridor is a 35-town region encompassing much of eastern Connecticut and a portion of south-central Massachusetts. The early stages of this partnership were previously reported in the Journal of Extension (Godin & Broderick, 2001). At that time the partnership revolved around a shared " Corridor Circuit Rider (CCR)" educator position and a new educational program in natural resources and land use for local decision makers. Now, 4 years later, the partnership is fully established under the name of the Green Valley Institute (GVI). It has grown to include 2 Co-Directors (one is the CCR) and five additional full and part time employees. Our audiences are targeted and specific:

1. Municipal land use commissioners;

2. Private farm and forest owners; and

3. Realtors, building contractors, and engineers who design and implement development projects.

The GVI continues to rely on data gathered in a previously reported needs assessment survey for program direction (Godin & Broderick, 2001). Traditional outreach vehicles such as workshops and print media were emphasized initially. For example, a 2-hour workshop module, Development Alternatives in the Quinebaug-Shetucket Heritage Corridor, was developed to address the top-rated issue in that survey. GVI now utilizes less traditional outreach tools as well, including a Web site and a list serve.

Digital Video and Web-Based Education

In early 2002, the GVI Extension educators began focusing on the Internet as a vehicle for reaching a greater percentage of our target audience. Research has shown that there are more than 97 million e-mail users in the United States, and 88 million Web users (Factoid, 2001), who utilize the Web to satisfy a myriad of informational needs. In fact, the Center for Media Research reports that most Americans expect to locate key information on-line: 60% already have access to the Internet and 40% of those surveyed have been online for three years (MediaPost Communication, 2003).

In response to this opportunity, GVI Extension educators created a multi-faceted Home page <http://thelastgreenvalley.org/gvi/> built around educational "Centers" for each specific target audience (commissioners, landowners, realtors). The site offers viewers a range of educational opportunities, from a tutorial on conducting a municipal natural resource inventory to a decision tree for landowners interested in protecting family lands from development. Building a credible Internet presence, however, was clearly only "half the battle." With the billions of home pages existing online, creative, attention-getting tools for attracting our audiences to the site were clearly required. We decided to utilize digital video in the form of a mini CD-Rom, direct mailed to the target audience, to acquaint them with our new communication channel.

Combining video and direct mail to reach a target audience is not a new concept. It's been used as a way to reach audiences with messages since the early 1980's (Tyson & Snyder, 1999). Research has shown that 90% of people that receive videos cassettes in the mail do watch them, and the conversion rates for videocassette direct marketing are as high as 23% (Nakamura, 2000). But the format in which video footage is reaching recipients is changing. CDs are now becoming more popular, and these mailings are being used in concert with a company or agency's on-line presence to connect with an audience.

CD-ROM direct mail pieces can bring an Extension educational message to life by communicating detailed information, including emotionally provoking video. It's a way to reach out to a target audience, grab their attention and then link them directly to a specific Web site--all the while making them feel involved in the process. One study, conducted by the Wharton School of Business, suggested that this method provides a response rate that is 600% higher than printed direct mail pieces (Nakamura, 2000). Little other empirical evidence can be found in the literature, however, concerning the efficacy of such campaigns in getting an audience to a Web site and subsequently influencing their knowledge levels and motivations.

The mini-CD we developed was co-produced by GVI and The Allen-Roche Group, a marketing/production firm from Boston, MA. It cost $13,500 to produce and print the CD and to print the CD mailers (Figure 1). The mailer and CD were designed so that the cost of each mailing was a standard postage stamp, or 37 cents. The CD itself contains about 6 minutes of audio and video, including interviews with GVI staff and target audience members and attractive aerial footage of the Heritage Corridor. It focuses on why the region is special, current threats such as land fragmentation and unplanned development, and how the viewer can get personally involved in solutions. Upon completion the CD invokes the viewer's Web browser and leaves him or her at our Web site. Our Extension education goals were to:

• Generate awareness and interest in GVI's programs,

• Get target audience members to view and learn from the Web site, and

• Motivate them to contact GVI and continue gaining and applying land use knowledge.

Figure 1.
Cover of the Custom-Designed Self-Mail Package for the Mini-CD-Rom

This article reports on the initial effectiveness of the direct mailed CD-Rom in accomplishing the above-listed goals. It attempts to fill a gap in the literature by providing empirical evidence for the effectiveness of a CD-ROM direct mail effort linking viewers with a Web site. Extension professionals who are interested in drawing attention to their own programs may find this outreach approach applicable to efforts involving specific target audiences.

Research Methods

The study reported here was launched in the spring of 2003. The Center for Social Research at Central Connecticut State University was contracted by the GVI to take the lead in designing and conducting the evaluation. The study utilized a quasi-experimental design with random assignment to either a treatment or control group. A comprehensive mailing list of the entire population of one target audience, municipal officials and land use commissioners from the 35-town region, was used as the sample frame.

The 35 towns represented by the list were stratified according to population density. Density ranged from 23 to 1,297 people per square mile. Towns were ranked lowest to highest, and all persons in every other town on the list received the CD. After deleting individuals who had moved or were deceased, the final count was 1032--518 in the treatment group and 514 in the control group.

The survey used a combination of mail and phone data collection methods. The mail component consisted of an initial first-class posting of the CD on April 4, 2003 to members of the treatment group. This was followed 5 weeks later by a postcard alerting members of both groups about the upcoming survey. Two weeks after this, questionnaires, cover letters, and postage-paid return envelopes were sent out.

There were separate versions of the questionnaire for each group. Each was designed by a graphics professional, printed in color, and took approximately 12 minutes to complete. Both questionnaires asked about a) familiarity and knowledge of GVI's mission; b) past, present, and intended future contact with the GVI; c) the usefulness of any previous contact; d) attitudes toward land use planning; and e) basic demographic-type data. The questionnaires for the treatment group included questions about the CD.

Response was tracked using code numbers affixed to the questionnaires. Three weeks after the first questionnaire mailing, a second identical mailing was made to nonrespondents. In addition, phone numbers were identified for 74.4% of the mailing list, and calls were made to these individuals requesting their cooperation. Five weeks after the second questionnaire mailing, the response rate was calculated to be 15.3%.

Approximately 3 weeks after the second questionnaire mailing, because of the low response, a phone survey was launched using a shortened version of the two questionnaires. The survey took 5 minutes to complete and was conducted by a university student trained in interviewing methods. All nonrespondents with obtainable telephone numbers were called once. The phone survey ran for 2 weeks, and a 40.6% response rate was obtained. The two principal reasons stated by respondents to the phone survey for not completing the mail survey were that they simply forgot (35.2%) and that they lacked the time (23.9%). Overall, 40.9% of individuals on the original mailing list responded to either the mail or phone survey.

Equivalency between treatment and control groups is an important factor in such studies if valid comparisons are to be drawn across groups and conclusions are to be made about the effectiveness of the experimental treatment. Fortunately, little deviation was found between the groups in this study. The treatment group made up 49% of the final sample; the control group made up 51%. Only three towns represented by the treatment group and four towns represented by the control group were not heard from, resulting in a very slight bias in the treatment group toward towns with greater population densities.

In their municipal capacity, 85.7% of treatment group members versus 91.8% of control group members were volunteers (versus paid staff). Concerning levels of formal education, 68.2% of treatment group members had a bachelors degree or above versus 77.5% of control group members. The sex and age of respondents was nearly identical for both groups. Approximately three quarters were men, and their average age was 54. When asked about the amount of nonformal training they had in a) land use planning and b) economic development (on a 1 to 5 scale, low to high), the responses for the treatment group were 2.12 and 1.71 versus 2.58 and 1.79 for the control group (respectively).

Findings

Before and after visits to the Web site were used as an initial indicator of the CD's impact. We collected data on "visits" (all "hits" to the main page caused by one user during a single logged-on session) rather than the "hits" themselves. Because a single visitor who moves around the Web site may return to the main page numerous times, they can generate multiple "hits" in a single session, thus distorting the numbers upward.

The Web site information was obtained from three different sources (Bravenet, netiQ, and ipowerweb), each of which counted visits to the GVI homepage (the page to which the CD connects a viewer). In the 2 months prior to the CD mailing, February and March, there were a total of 122 visits to the main page. In the 2 months following the mailing, April and May, there were 324 visits, constituting a before/after increase of 265%.

"Directory visits," i.e., visits to any GVI sub-pages listed in the main page directory, were also counted. These visits rose steadily from 422 in April to 1,006 in August. Interestingly, the number of main page visits actually declined slightly during this same period, suggesting that repeat viewers were bookmarking sub-pages of interest and returning directly to them for more detailed analysis and use.

Survey analysis revealed that 60% (n=123) of treatment group respondents reported having viewed the CD at least "moderately." Others in the treatment group reported viewing it "slightly" or not at all. To assess the effects of the CD, moderate viewers were compared with a combined group of slight/nonviewers from the treatment group plus members of the control group.

Use of the Web site by moderate to high viewers of the CD increased dramatically after viewing the CD, increasing from (mean score) 0.49 to 1.48 on a 0 to 4 scale (not at all to one hour or more). In addition, moderate to high viewers of the CD were significantly more apt than slight/nonviewers to display the characteristics listed in Table 1.

Table 1.
Comparison of Moderate to High and Slight/Nonviewers

	Mean Score Mod to High Viewers	Mean Score Slight/ Nonviewers
Be more familiar with activities of the GVI (1 to 5 scale, not familiar to exceedingly fam.)	2.83 (n=123)	2.50 (n=298)	t=2.65 p=.008
Have viewed the Web site more (0 to 4 scale, not at all to one hour or more)	1.48 (n=123)	1.18 (n=298)	t=2.17 p=.031
Think the Web site was more useful (0 to 5 scale, not applicable to exceedingly useful)	2.33 (n=120)	1.67 (n=285)	t=3.94 p=.000
Have greater intentions to phone GVI in the future (0 to 3 scale, no intention to definite int.)	1.43 (n=35)	0.91 (n=116)	t= 2.74 p=.007
Have greater intentions to e-mail GVI in the future (0 to 3 scale, no intention to definite int.)	1.20 (n=35)	0.75 (n=116)	t= 2.41 p=.017

The average response for those who received the CD but viewed it slightly or not at all (n=82), when asked to what degree they intended to view the CD in the future (on a 0 to 3 scale, no intention to definite intention) was 1.39, meaning there was a slight to moderate chance they would. When 50 individuals from this group responded to a question asking why they had not had a chance to view the CD more thoroughly, the most commonly mentioned reasons were a) that they had not had enough time or interest (mentioned by 58% of respondents) and b) the fact that they did not remember receiving it (mentioned by 22% of respondents). The latter reason may be an indication that it was treated as junk mail.

When asked to list what was most useful about the Web site, the most frequent responses from members of both treatment and control groups who had viewed the Web site more than 15 minutes (n=184) was that it was a) good source of objective information and b) quick access to information (though slightly less than 10% chose to answer this question).

The average response from members of both treatment and control groups who had viewed the Web site less than 15 minutes or not at all (n=237), when asked to what degree they intended to visit the Web site in the future (on a 0 to 3 scale, no intention to definite intention) was 1.30, meaning there was a slight to moderate chance they would. When 103 individuals from this group responded to a question asking why they had not had a chance to view the Web site more thoroughly, the most commonly mentioned excuses were a) that they had not had enough time or interest (mentioned by 67% of respondents) and b) the fact that they did not have a computer (mentioned by 15% of respondents).

It is interesting to note that not having a computer was only mentioned by 8% of those who viewed the CD slightly/not at all and 15% of those who viewed the Web site less than 15 minutes or not at all--a reflection of how common computer ownership is now, even in small rural towns. When all respondents (both treatment and control groups) were asked to rank their preferred means of learning about land use planning and/or sustainable economic development issues, they mentioned (in priority order, using mean scores, on a 1 to 6 scale, most to least favored): publications (4.72), workshops (4.58), Web site (3.67), video (3.14), email (2.87), and phone (2.64).

Conclusions

The study reported here provides empirical evidence for the effectiveness of a CD-ROM direct mail effort linking viewers with a Web site. Findings show that an organization such as Extension can expect more than half of those who receive a CD in the mail (60% in this case) to view it at least moderately and that these individuals are significantly more apt because of this activity to:

• Have greater familiarity with the organization's goals,

• View the Web site more,

• Think the Web site is more useful, and

• Have greater intentions to contact the organization for additional information/assistance in the future.

This study shows that a CD-Rom direct mailing can have a significant impact on the awareness of an Extension program such as the Green Valley Institute, by a target audience such as municipal commissioners and large landowners.

It is noteworthy that for the type of audience in this study lack of computer equipment was not much of an impediment to using "new" communication technologies. Yet this type of audience still prefers more traditional Extension learning channels (e.g., publications and workshops), though Web -based communication is not far behind in terms of preference.

One caveat to using direct mail CDs should be pointed out. As this technology becomes more common, there is a chance that growing numbers of people may at first glance consider it junk mail and deal with it accordingly. Forty percent of the CD recipients in this study stated they viewed it slightly or not at all, and when a sample of these individuals were asked, 22% said they could not remember receiving the CD in the mail. This "junk mail" phenomenon will need to be included as a cost of doing this kind of outreach.

References

Bravenet. http://www.bravenet.com

Factoid. (2001, January 4). EDN.

Godin, K., & Broderick, S. (2001). Partnering with a national heritage corridor: A Connecticut case study. Journal of Extension [On-line], 39(5). Available at: http://www.joe.org/joe/2001october/a3.html

Ipowerweb. http://www.ipowerweb.com

Nakamura, B. (2000, July 10). If you want your marketing to be profitable...read this. Brandweek.

NetiQ. http://www.webtrends.com

MediaPost Communications. (2003, January 8). The Internet is a mainstream information tool for 97 percent of all users. Retrieved February 3, 2003, from http://www.mediapost.com

Tyson, B., & Snyder, L. (1999). The impact of direct mail video. Public Relations Quarterly, 28-32.