Twitter IS A Broadcast Medium, NOT A Conversational One And Implications For Customer Service

Twitter IS A Broadcast Medium, NOT A Conversational One And Implications For Customer Service

Over the last two years, I’ve noticed that conversations (dialogue, responses to tweets) have been harder and harder to find despite that fact that social media mavens keep talking about Twitter as about relationships, interaction and engagement (from a business perspective). It’s really noticeable.

So, I decided to ask people where I could find the interaction that so many people think happens on Twitter. Where IS the dialogue? Where IS the great conversations.

I asked for people to submit hash tags I could look at, and from which I could collect data to verify conversation and interaction were taking place. I also asked that suggestions NOT include hash tags related to social media, or chat hash tags, since I am interested in how the average person, not social media marketers, are using Twitter.

Here’s some preliminary results from a limited sample size, and from the one hash tag I’ve looked at. I’ll continue to update and re-sample.

  • 30% of tweets were retweets
  • 17% were thank you’s
  • 40% were one way broadcast of links, marketing, etc.
  • 9% were autotweets, just repeating the same message but with the recipients changes.
  • At BEST 3% of the tweets sampled were replies (i.e. conversations)


It shouldn’t be. It’s all out there in front of all of us, but what IS remarkable is that the person who submitted the hash tag is a smart guy who I respect. But he didn’t notice.

It shouldn’t be shocking if you consider the nature of Twitter, something I’ve talked about repeatedly. The characteristics of Twitter simply do NOT encourage conversation.

First, it’s a streaming medium, which means that the “life” of a tweet (the amount of time it might garner responses) is measured in hours, and not many of them. You post. People respond if they see it, and then you won’t get many responses after an hour or two. Hence conversation is close to impossible overall.

Second, of course is the character limit. It’s very difficult to “converse” in 140 characters. You can actually test this out on the phone by trying to have a conversation when each side can only use eight words per “turn”.

Implications For Customer Service

To make use of Twitter for customer service, we need to know how people ACTUALLY use Twitter, and put aside how “experts” think people SHOULD use Twitter.

Putting aside the issues of sample size, and the lack of ability to generalize from our puny data (see below), if we assume that most people do NOT interact on Twitter (dialogue, conversation) what does it mean?

Quite simply it means that people are not behaving as if they are interested in dialogue with you. It means that the likelihood that you can address customer concerns (we’re talking customer service here, not marketing) on Twitter alone is virtually zero.

You will need to shunt Twitter customers to another channel, so that actually puts additional barriers between you and the customer, and INCREASES time to solution.

It means you can’t form real relationships with your customers solely on Twitter. At best you can only form “thin relationships”.

There’s many other spin-off implications that are POSSIBLE, but more speculative. What we DO NOT know is how people read tweets, digest content, or otherwise process the information in tweets. While I don’t have direct data, I’d suspect that do to the thin content and thin relationships, most people wouldn’t remember 99.5% of tweets an hour after reading. That they skim at best, and that the tweets that are remembered are those from real world friends and family, and NOT your tweets (on average).

Your Challenge:

I challenge anybody reading this to go to their favorite hash tag and classify the tweets as I have. Post your results, and if nothing else, you’ll be helping me find intelligent engaged life on Twitter.

Author: Robert Bacal

Leave a Reply