Reading tweets: it’s complicated….

Conversational media may be new to brands, and talk may make for a new kind of marketing, but there’s little new to what we talk about, do with it, do it for, and with whom. What’s changed is the medium we use, and with that, how we phrase our talking.

Even in image and eyeball branding, marketers and advertisers have known how to approach tone and voice in their messaging. Pictures tell a thousand words, and they too express subtleties and nuances that can be used to strategically appeal to and target audiences.

The differences with conversational media, of course, are that the medium is itself personal, and communication is two-way. Brands now reach audiences on personal media and in personal accounts, both of which make two-way conversation possible. Much of this talk also happens in public, or in front of audiences (followers, fans, etc), so there is an additional constraint on one’s style, manner, and content of “speaking,” for much of it will be overheard (and searched, tracked, etc).


When we hear of brand “voice,” we might think generally of a brand’s use of twitter, and more specifically of how it talks. How personal is the brand. How friendly, responsive, helpful. Or how authoritative, useful and informative. These are not all voice, technically speaking, though I don’t mind if we simplify matters occasionally by referring to a brand’s “voice.”

There are things we say, how we say them, and what we do (or is done) by saying them. Each statement (tweet) may also solicit or attract a response. For example, tweets can be informative, in which case they may be passed around for their information value. If they come from an authoritative source, they may be passed around in part because they have been spoken by an authority.

Take @guykawasaki, whose influence rank is based on his follower count, activity, and retweet count. @Guykawasaki is good for a lot of industry news, whether you agree with him or not. He is a news service for many twitter professionals. His social capital is built less on his personal opinions and more on his audience reach value as a source of news. One might say that his capital, and that of other individuals who use their accounts as a news wire of sorts, comprises the twin values of news: information and timeliness. (The value of news is that it is new. News actually means two things: the content is relevant, and new. Old news is less valuable not because it is false or lacking information, but simply because it is no longer new. In fact the persistence of information that was once news transforms it over time into something else: a story.)

@Guykawasaki may get a lot of retweets, but some will be simple rebroadcasts, and some will also confirm and contribute to his status. A smaller number of retweets may serve as agreement with his opinions, but only a small number. The social capital of a newsmaker is not expert capital, because expertise requires opinions, perspectives, and analyses. (This is not to say that Guy Kawasaki is not an expert — I’m speaking here only of tweeting activity)

By contrast, @jowyang, whose rank is also up there, tends to tweet his own perspectives, activity, and news. He voices as an analyst (which he is). When he is retweeted, it is more likely that a user confirms his authority as analyst implicitly or explicitly. When @jowyang tweets a perspective, retweets implicitly affirm that perspective. Retweeting, in this sense, comprises a kind of conversation by proxy: the retweet is a repeat or citation, and in many cases nothing is added to the retweet to reflect on or add to the original tweet. The retweeter can associate the perspectives offered in the original tweet with his or her own, thus speaking by proxy. And possibly attributing proxy status of the original tweeter to oneself.

(There is a strange reference to the social operation of mimesis here. In mimesis, symbolic acts and values are reproduced within cultural practices by means of imitation. To imitate or copy cultural influences is in some ways to seek the status of the influence(r) by means of imitating him/her. People do this stylistically, or by behavior, activity, character, and so on. Retweeting, in some cases, may also involve this kind of social logic, but operating by “stealing” a person’s words. Not putting words in another’s mouth but taking them out… The practice of retweeting in a normal conversational setting would appear bizarre indeed, not to mention make a farce of conversation! Conversational media are in many ways socially dysfunctional.)

These two examples involve “influencers” who have also established themselves as personal brands. Corporate brands are for the most part a different case altogether. In their case it is not a matter of branding the person, but personalizing the brand. Hence the importance of face, personality, character, and behavior. Using conversational media, the brand can make use of the opportunity to build relationships, humanize a reputation, reshape image and impressions of that image, and of course actually provide service, support, help, and other kinds of direct customer assistance. The fact that all of this occurs for posterity (search), and publicly, provides an abundance of strategies, tactics, and techniques.

Because the means of production is also a means of distribution (communication technologies), conversational branding involves matters germane to communication as well as media. In communication is the distinction between what is said and how it is said, as well as the performative (what is accomplished in the saying of it. For example, help is actually provided, politely, by means of providing information requested by a customer. That’s the communication. In terms of media, the fact that this service can be seen by others contributes to perceptions related to reputation, authenticity, “image,” responsiveness and so on.

We could systematically go through the effects and value of different kinds of statements. In the interest of keeping this short, however, I’ll simply offer some types of statements and we can perhaps pick them up in comments or future posts. Note that these statements include different kinds of content as well as different expressions:

Some types of statements

  • News and announcements
  • Information
  • Invitations
  • Questions
  • Answers
  • Recommendations
  • Offers
  • References (links to videos, podcasts, pages, products, events, etc)

Each of the above statement types does something, appeals in some way to a response or action, reflects on its author, and may count towards relational value. (Sometimes called social capital. I say relational because social capital is not a fixed attributed, thinglike, owned or possessed, but is instead a measure of (power/influence) relations. Social capital has no “existence”: it is actualized, realized, put to use only by means of interaction/communication with those whose perception of that capital value determine its influence. It is a kind of power belonging not to its owner, but to those who believe in it.)

The differences between these kinds of statements (tweets) not only relate to what their author finds interesting or worth tweeting about. They also suggest responses (retweets, @replies, follows, and conversation). Questions, for example, suggest what counts as an answer. Recommendations indicate taste. Statements of location may serve as an invitation to meet (they are ambiguous). Invitations are easily passed along. All of these kinds of statements suggest responses or actions that others can take around the message. They are just a small number of examples of how things get done by means of talk — and how it is that conversation by means of tools like twitter must compensate for the fact that messages are out of context, situation, and face-to-face dynamics.

Some of these twitter-based acts of conversation accomplish or achieve something for followers, or for certain followers. Some help. Some answer. Some inform, and so on. But tweets are seen by a public — and this may be reflected in how they are written. Users seem to vary in their sense of the twitterverse as populated by people, whether those people are friends, and whether they’re paying attention. Users vary too in their relation to the medium, which may be a personal communication tool, a social tool, a community service, or public medium. These differences in proximity and relationship to the medium and its audience are broad; and they make it very difficult to identify strategies or approaches for twitter.

There is no simple correlation between what a user or brand does, and how it should “voice” itself. Every tweet, regardless of its author’s intentions, will have meaning attributed or interpreted by those who read it. Meaning in social technologies is less in the hands of the author and more in the hands of the reader, for there’s less of an opportunity to negotiate or work out a common understanding. A brand’s voice on twitter is not entirely in its control: audiences draw their own conclusions.


There is much more that could be said on this, time permitting. But I’d rather continue this in comments. There are too many strategy and tactical differences and nuances to cover properly in a single blog post.

This entry was posted in Twitter and tagged , , . Bookmark the permalink.
  • Tamsen McMahon (@tamadear)

    Have you seen ? I'd be curious about (a) what you think about Rajesh's construct and (b) how you'd map your tweet types to how people read them.

  • gravity7

    I hadn't seen lateraction — thanks for that. It's an interesting take — unscientific as is mine ;-)

    I would want to add in factors that are not specific to content or message, but which might relate to the brand (of blog), reputation (author), network, comments, and so on. Which I think is fairly obvious. Content alone isn't a reason to follow a blogger.

    Knowing that the blogger has influence, is first as a newsmaker, is a kingmaker among peers, is an evangelist with access to sneak peeks, and so on — these are reasons to follow a person but which are not “in” the content per se.

    So yes, I'd map tweets to personalities and behaviors, and would look for these kinds of distinctions. I'd be careful to distinguish the conversationality of twitter from the publishing of blogging. Twitter's not micro-blogging to me, but is talk. So I'd look at the content of messages differently for that reason also.