Tuesday, December 09, 2008

Forrester Report on Blogging & Trust

Forrester today released a new report on corporate blogging and trust (registration required, but free). I'm glad to see they ask, "Is a corporate blog worth doing?" Rather than just assuming that everyone should be blogging, Josh Bernoff and his co-authors actually acknowledge that some blogs are not worth the effort. Of course, that's a hard point to ignore when they report that only 16% in the survey said they trusted company blogs.

Among the tips offered:
"Honest and transparent blogs will get noticed. Those who write in a corporate voice will be ignored and ineffective. What types of blogs will consumers trust? Those that reveal tidbits about what’s going on inside the company, those that comment intelligently on customer problems and competitor products, and those that speak like people. Robert Scoble pioneered this technique at Microsoft years ago, but it’s still hard for companies to figure out."
Thanks to Steve Rubel <@steverubel> for tweeting this and Mihaela Vorvoreanu<@prprof_mv> for re-tweeting.

4 comments:

Daniel C (Dan) Smith said...

Thanks for the link -- it pushed me to download the report. I already had seen Josh's post as I subscribe to the Groundswell blog but I did not download the report. Individual bloggers may be more aware that corporate blogs take a big risk if they distort the truth much.
I have some related data with a much smaller sample that employee social media use inside and outside work correlates with trust in fellow employees.
Here is a study in the peer-reviewed literature that found a link between internet use and interpersonal trust:
Beaudoin, Christopher E. (2008) Explaining the relationship between Internet use and interpersonal trust: taking into account motivation and information overload. Journal of Computer-Mediated Communication. 13 550-568.

Anonymous said...

Let me start by quoting an important fragment from the methodology section (as posted by the study's author):

"A note about how we collect data. The data comes from an online survey we conducted in Q2 of this year. Our online panel is as representative as we can make if of the US online adult population (18 and older). [...] Respondents could also answer that they didn’t use a particular information source. In this case about 80% of those we polled said they did use corporate blogs. Of those who used them, only 16% rated them 4 or 5 on the five-point trust scale."

What mystifies me is how many people responded that they "use" a company blog - 80%. Really? How credible is that when you're considering the general population's consumption of blogs? And if it's not, and the online panel used for this survey is not representative for the general US online population, then why do we even discuss about the 16% of them that trust company blogs?

And what does it mean, anyway, to trust such a blog? It's unclear what, exactly, a "company blog" is (for example, are NYTimes or Huffington Post blogs "company blogs"?), what does it mean to "use" them (how frequently are the respondents reading these blogs?), and what the 16% number represents (since the trust was measured on a 1 to 5 scale, 1 = 'don't trust at all').

Also, it's kind of ironic that the report gives advice on what to do in order to increase company blogs' credibility, since the survey didn't probe in any what are the factors that determined the existing level of credibility.

What's your take on the study? :)

Tom Kelleher said...

Thanks for the link to more info on methods, Constantin. I tried to check that out myself by following the footnotes and ran into a $379 roadblock!

In general, I take the representativeness of the sample with a grain of salt. But I think the general idea that many people who reported "using" company blogs also are reluctant to say they trust them is still worth noting, especially when presented by someone so enthusiastic about the potential of blogging for "ROI." In this case, a general-population sample may actually be less appropriate. Asking people who say they don't "use" blogs to report how much they trust blogs might be an interesting intellectual exercise, but I'm not sure that's what this study was about.

I don't think that the sampling issues undermine the take-home point that there's more to earning trust than *mere* blogging behavior.

That said, your point on methods is well-taken, and shows the need for more academic research to corroborate (or not) what the pros are reporting.

As for the trust concept, Dan Smith (who comments above) is taking a close look at that, though more from an internal org-com view. He's now working on the hard part -- getting good data. So if you know anyone working in a big org who could help him survey employees on blogging and trust, you could help us balance some of the trade lit with theoretical work.

Anonymous said...

Tom, I'm going to send you the report, it's actually free with registration, I think.

I see your point about the significance of people's reluctance of saying they trust company blogs. It's just that there's no way to say how much the wording of the question affected the response, and as a consequence the 16% number is meaningless.

Just to clarify: I don't have a problem with that number - it's fine with me if only 3% of people think company blogs are credible, as long as the research is correctly done. But it looks to me that the instrument use for this survey has serious validity problems.

You're right, there needs to be more academic research to validate (or not) the pros' reporting, but I would think that Forrester's work aims for highest standards. After all, Forrester is a research company, which brings a lot of responsibility, since the results of its studies are reported accordingly (just look all over the blogosphere -and even mainstream media outlets- today to see how many people are breathlessly reporting the 16% figure, without any word of warning about methodology concerns).

For other people's take on this check out the discussion between Max Kalehoff and Pete Blackshaw, as well as Peter Kim's response.

(Thank you for the pointer on Dan Smith's research, I'll check it out.)