Case Studies:
Consumer Packaged Goods

Consumer Package Goods Case Study: "Company A"

Objective / Brief

Company A is a leading developer, manufacturer, and supplier of printing solutions for offices and homes. Since 1995, Company A has analyzed new forms of electronic media (e.g. Websites, newsgroups and chat rooms) to identify issues with its products that could negatively impact its brand reputation and to better understand the relationship between traditional media and consumer opinion, both of which are often expressed online. In 2006, Company A tasked KDPaine & Partners (KDPP) to update its analysis of consumer-generated media (CGM), which in 2003-2005 was focused on newsgroups, in order to:

• Identify and track trends in consumer sentiment about inkjet printers and supplies.

• Relate these trends to traditional media coverage.

• Provide recommendations to Company A about engagement with consumer-generated media.

• Achieve each of these objectives in a manner that maintains the validity of the investigation to communicators, despite recent shifts in the CGM environment.

Strategy

KDPP and Company A first looked to commercial and academic research on the growth of the blogosphere to evaluate whether a proposed supplantation of the newsgroup investigation with one focused on blogs was grounded in empirical data. Among the findings, by the end of FY2005, blogs had grown to a population of 53.4 million and counted 27% of American Internet users as readers, which included 41% and 30% of the desirable Gen Y and Gen X markets, respectively. This was compared to a last-estimated population of 190,000 public Usenet groups , in which no more than 10% of online Americans participated. Research by Columbia University and Euro RSCG Magnet had also shown that 51% of journalists were reading blogs regularly, with 28% relying on blogs for day-to-day reporting. (See Exhibit A for endnotes).

KDPP and Company A agreed that an updated CGM investigation was necessary. In addition to blogs, KDPP and Company A recommended the inclusion of discussion groups, Internet forums, message boards and social networking sites, which also provide insight into consumer opinion via first-person writing, but have user cultures and operations that are distinct from blogs. To enable comparison with the results of Company A's ongoing traditional media measurement program, KDPP recommended using Company A's current measurement instrument, with a priori adjustments to measure the medium-specific variables of post part (body, comment), genre, commenter, and interconnectivity (measured by number of hyperlinks and number of trackbacks). Human coders were also trained to identify emergent variables. As inkjet printers and consumables represented XX% of Company A's total sales in 2005, these product categories were the primary subject of the 2006 investigation.

Execution / Implementation

To accommodate latent measures of tonality, issue positioning, and message communication, and to more accurately account for the sarcasm, humor and irony previously observed in CGM, KDPP did not employ any natural language processing tools in this analysis, relying instead on a team of human coders.

Beginning in FY2006 (April 1), each business day, a collection team would extract all posts and comments relevant to the inkjet printers and consumables produced by Company A and its tracked competitors using free-text searches on Technorati (which indexed 27.2 million blogs in 2006), BlogPulse, Google BlogSearch, Google and IceRocket (which includes comments in its indexed pages). Relevant content was uploaded and coded via KDPP's online coding module, to reduce errors in data entry from coding sheets. Spam blogs, spam sites and made-for-adwords (MFA) sites were removed. Content produced on weekends was collected the following Monday. All English-language content was included, regardless of author nationality, as American readers would have an equal opportunity to view this content online.

870 sampling units (posts, messages, comments) were extracted from 531 sources. 67% of discussion mentioned Company A.

To protect ecological validity, coders analyzed content directly from the source page, not via text archived in KDPP databases, which is uploaded only for historical purposes. This facilitated the coding of outbound hyperlinks, inbound trackbacks and images that might otherwise be lost, and avoided the analysis of advertisements, which are often collected by vendors that automatically gather Internet content.

With the release of search-based RSS feeds by Technorati in September 2006, KDPP employed this method of content collection as a source for Technorati-indexed blogs, which was validated at the end of each cycle by free-text searches, to ensure as robust a sample as possible.

To keep the analysis cost-effective, full reports were delivered to Company A quarterly. To keep the analysis timely and relevant, KDPP also provided Company A with real-time, dynamic charts and data extraction tools, both with full drill-down functionality, via a Web dashboard. At the end of FY2006, KDPP and Company A presented the findings of the CGM analysis in the annual presentation of traditional media findings to C-suite level executives of Company A and leaders of the Company A communications teams.n.b. Research teams from both KDPP and Company A acknowledge the limitations of the non-random convenience sample employed in this investigation, but as a population of CGM content from which to randomly sample has yet to be compiled by any investigator, this was considered acceptable. Within Company A, generalizations drawn from the CGM study are noted to reflect this limitation.

Conclusions

A positive correlation existed between top-tier traditional media and consumer-generated media. For FY06, the volume of traditional media and CGM content published about Company A had a weak, positive correlation (r=.51), which, in light of Company A's non-engagement with CGM, suggested either two independent media acting in concert without the influence of public relations practitioners or a weak intermedia influence. When traditional media content was restricted to product news/ reviews, the correlation becomes stronger (r=.64). The relationship was especially true in September, when, as part of its holiday outreach, Company A's sent out more review units to traditional journalists than it had in any previous year. This resulted in a 78% increase in product news from August to September in top-tier traditional media and a 526% increase in CGM discussion. Company A's 2006 holiday sales also exceeded expectations.1

In the 2007 holiday season, Company A has been awarded additional review models, with the expectation to further increase exposure in the traditional and consumer-generated media, in the hopes of a positive impact on sales. Company A's decision in August 2006 to not distribute review models to bloggers was also supported by these findings. CGM discussion still increased in September (and maintained its momentum well into October, when traditional media coverage had subsided) and Company A avoided the risk of blogger review model programs, which have recently brought criticism to companies like Microsoft and Nikon.

CGM authors did not take their queue from traditional journalists as to tonality, but agreed on key product benefits. 33% of Company A's CGM content in FY2006 was considered positive toward the company, compared to 49% of traditional media content. Of the five most-discussed favorable issues for Company A's products in CGM (Quality of Output, Ease of Use, Speed, Versatility and User Benefit), Quality of Output, Ease of Use and User Benefit were also among those highlighted by the traditional media, which focused on Price and Breadth of Capabilities, as well. There was a larger disconnect in unfavorable issue positioning: CGM focused more on Customer Service, Reliability and Ease of Use than did traditional media, which was more Price and Speed-focused. These findings suggest that traditional and consumer-generated media may agree on the benefits of a product, but diverge as to the factors that can interfere with satisfaction (namely, service vs. price).

Exhibit A: Research Endnotes

1 To guard impartiality, KDPP was not provided with any sales data or information about PR outreach efforts until after the analysis was completed.

I Henning, J. (2005, April). The blogging geyser. Perseus Development Corporation White Papers. Retrieved April 30, 2006, from http://www.perseus.com/blogsurvey/geyser.html.
II The Pew Internet and American Life Project. (2006, January 22). Generations Online. Retrieved March 1, 2006, from http:/pewinternet.org/pdfs/PIP_Generations_Memo.pdf.
III Joyce, E., and Kraut, R. E. (2006). Predicting continued participation in newsgroups. Journal of Computer-Mediated Communication, 11(3).
IV Ridings, L., and Gefen, D. (2004). Virtual Community Attraction: Why People Hang Out Online. Journal of Computer-Mediated Communication, 10(1).
V Euro RSCG Magnet. (2005, June 20). Great Thoughts: Turning Information Into Knowledge. Retrieved April 30
2006, from http://www.magnet.com/index.php?s=_thought. 

Five reasons why KDPaine & Partners should be measuring your success in social media or traditional public relations:

1. All our research is custom designed and performed by human experts: Sometimes computer-aided, always human-coded.
   Our team has been measuring and evaluating communications for 25 years.
2. We measure the tough stuff!
   It's not just about media relations anymore. We measure image, public relationships, reputation, outcomes, and social media conversations.
3. We don't just give you numbers -- we give you advice.
   We are your partners in continuously improving your programs.
4. Scaleable research and services to meet any budget.
   Our solutions start at $3,600.
5. Proven science, proven techniques, proven results.
   We adhere to the Institute for Public Relations Guidelines for Measuring Public Relations. We never provide Advertising Value Equivalency or any numbers based on inexact science.

Click here to purchase the newly released
Social Media Measurement White Paper

25 Women That Rock Social Media

8th Annual Summit on Measurement
Portsmouth, NH
October 6-8, 2010