In 1966, Harvard Business Review published a study by Ernest Dichter that identified four motivations for a person to communicate about brands, as follows:
- In 33% of the cases, people shared because of product-involvement. The experience was so novel and pleasurable that it had to be shared.
- In a quarter of cases, people shared because of self-involvement. Sharing knowledge or opinions was a way to gain attention, show connoisseurship, feel like a pioneer, have inside information, seek confirmation of a person’s own judgment, or assert superiority.
- One-fifth of sharing occurred from other-involvement: the speaker wanted to reach out and help to express neighborliness, caring, and friendship.
- In another 20% of cases, message-involvement drove sharing. The message was so humorous or informative that it deserved sharing.
Implications for Brands
If we assume those trends to be valid and normal expectations for online sharing today, we can use that distribution as a benchmark to assess the health of the conversation about any brand, based upon the extent to which the conversation about the brand deviates from this average distribution.
For example, we might expect 33% of conversation about a brand to focus on people describing their experience with the brand. And when we see more than 33% of a brand’s conversation being generated by customers based on product or service experience, we might guess that the brand is doing something exceptional — either within the customer experience, or within their social media marketing.
When we see a brand whose customers create significantly less than 33% of the conversation about the brand, we are likely to discover flaws in either the customer experience or the brand’s social media capabilities.
Category Benchmarks
In the apparel category, we have seen brands with more than 90% of their conversation generated by consumers talking about their experience with the product. On the other hand, we have also seen apparel brands with less than one-third of their conversation occurring based on customer experiences with the product.
For one automotive client, we found that 28% of consumer messages about the brand described their involvement with the product. That’s slightly below the 33% general benchmark, but Converseon analysts determined that 64% of the messages were positive. (No automated sentiment scoring was used.)
On the other hand, for one of our B2B technology clients, 32% of online customer messages describe the customer’s involvement with the product. However, only 24% of those messages were positive, as measured by Converseon analysts.
The difference, we found, is that the technology company receives a greater share of mentions within online forums, where customers are trying to solve problems. In general, sentiment in support forums is more negative than other venues, as indicated in our recent research with Wendy Moe of the University of Maryland, David Schweidel of the University of Wisconsin, and Chris Boudreaux of Converseon.
So, any brand that receives a lot of mentions in forums should expect that (1) more of their mentions describe customer experience with the product, and (2) more of their mentions will be negative.
On the other end of the spectrum, for one of our shampoo clients, 73% of online consumer messages described the consumer’s experience with the product. In addition, we noticed consumers in the shampoo category are far more likely to mention multiple brands in a single message, whether the brands are complimentary (e.g., hair spray and shampoo) or competing (e.g., 2 shampoo brands).
Average sentiment, as measured by Converseon analysts (not algorithms) for the shampoo brand was 65% positive — far greater than the auto or the tech brand. Most of the shampoo brand mentions existed within product reviews, recommendations and anecdotal messages such as status updates. And we know from Converseon research that those venues tend to be more positive, on average.
In general, brands need to understand the average sentiment for their category, which is something we benchmark for all of our clients. In addition, we provide benchmarks across industries, and we help brands identify specific root causes of any variances in their sentiment, to help a brand set reasonable expectations about the extent to which their sentiment might ever change.
Finally, brands need to understand the venues where their mentions occur, and realize that the mix of venues predetermined limitations on your overall sentiment. Therefore, sentiment must be measured within each venue type, over time.
In fact, we now have a model that corrects for such anomalies, and allows us to measure true Customer Satisfaction, with reliable correlation to traditional satisfaction surveys, using online conversation research.
Remember: shifts in sentiment toward your brand may simply result from the conversation mix shifting across venue types. Just because overall brand sentiment changes, that does not mean your customers feel differently about your brand.
If you are reporting overall brand sentiment to your executives, you really should stop.
And thanks to Will Bottinick and his research team for pulling together these numbers.
Chris Boudreaux leads the Business Integration practice at Converseon, where he helps leading brands achieve business objectives through social media by transforming business processes, data integration, and governance. Chris can be reached at via email at cboudreaux@converseon.com or via Twitter.