The Social Conversation Misconception – Who Stole Engagement And What The Lack Of Conversation Means For Customer Service
Advisors on social media are quick to point out that whether companies are part of the process or not, your customers are having conversations about you, the implication being that companies better get in the game, so they can influence where those conversations go, and how they are perceived. Of course, since social media is all about conversations, people talking WITH each other, the face value of this statement should be clear. Except it’s not true. If we consider a conversation as involving two or more people interacting (tweeting, commenting) back and forth; a dialogue, in effect, we find that there is remarkably little conversation going on. You’d think that since social media allows us the capability to have these dialogues, that they are taking place, but once you look at how real people use social media, all of a sudden a different picture emerges. When dialogues occur, they are remarkably short, hardly resembling conversation. In fact, social media is a “broadcast” medium, which means that it consists of a lot of people, one by one, just “sending stuff out” and not “engaging” with others. This is one of those situations where the capability and how real people take advantage (or not) of the capability are quite different.
It may be stating the obvious to say that if when you have a majority of tweets (or updates) that receive NO response it all, it’s not a “conversation”. It still needs to be said. Of the 30 percent or so of tweets that receive a response, (retweet, link click, @response), obviously the only one of those that involves conversation occurs for the @response. So, not only are some 70 percent of tweets ignored completely, but of the remaining bunch, many of them receive no “conversational” responses. When you consider that, there’s not much left over that falls into the “Hey, we’re talking WITH each other”.
More specifically, DUNCAN GEERE, writing on Wired Magazine, and using Sysomos research findings puts it this way:
… 23 percent of messages … get an @reply. Drilling down, Sysomos found that 85 percent of replied-to messages get just one reply, 10.7 percent get two, and just 1.53 percent get three replies
Don’t get confused by the multitude of numbers. Let’s simplify it and consolidate the statistics:
For every one hundred tweets:
- twenty three receive an @response
- About twenty receive ONE @response
- About two get two @responses
- Only .345 tweets get three responses
Understanding “The Stream” Explains Why Social Interaction Is So Limited
There are several reasons why so few tweets receive response, and there’s so little conversation happening on Twitter. The same reasons also apply to Facebook.
Twitter named its flow of tweet a “stream”. It’s actually a great metaphor. Imagine sitting on the edge of a stream, watching the water flow by. Most of the time you’ll be starting, probably blankly at the water that flows, not paying much attention, just taking it in in a rather inattentive way. Then something different might float down the stream, an animal carcass. BRIEFLY, you take note of the aberation. You might ponder the sad fate of the animal, or wonder about how the poor creature came to be going past you, but of course the carcass quickly floats away, out of sight, and unless you’ve had a big emotional reaction, it’s a case of “out of sight, out of mind”. If, while pondering the animal, a small yellow volkswagen is carried by the water past your vantage point, you will quickly attend to that. And on and on. What is in front of you is there only for a very short instant. In the next instant, something else comes along. And, if you hadn’t been sitting at that exact spot, at that exact time, you never would have seen the dead dear, or in fact, the yellow volkswagen.
Twitter is just like that. There’s a flow of tweets going past you, and you sit on the bank, either observing the movement in a vague way, until something catches your attention. You might retweet it. You might respond with an @response. In all likelihood you’ll do neither. The tweet flows by, replaced by the next and the next, and it’s highly unlikely you’ll even remember what you saw in the stream.
If you hadn’t been sitting watching the tweet flow at that exact time, you wouldn’t have seen the tweets sent at that time, and you’d probably NEVER see them.
We know that almost ALL responses of any sort to tweets occur within the first minutes or hours after it is sent. Research on it has produced slightly different results, but the maximum is about three hours. Then, it pretty much ceases to exist.
Tweets are disposable. They are time specific. Add to that that the huge majority of tweets aren’t terribly attention grabbing. Many are self-centered, or clearly attempts to market. Very few have any lasting impact, or stand out enough to provoke even a minimal response. When you think about it, it’s rather shocking that only about six percent of tweets are even retweeted, given how little effort is required to retweet. It’s pretty much one click, and that’s it. But only six percent are retweeted.
In case you are wondering about whether tweets or status updates can be “revived” because, at least in theory one can do a search for a particular topic, that doesn’t work either. You can prove it for yourself by doing like I did, and do a search for “customer service” using the Twitter interface. I let the search run for five minutes or so, scrolling down to encourage it to continue to find new customer service related tweet. And, boy, did it. There were so many, I couldn’t possibly count them all, but what I did learn was that it ordered tweets according to recency, and unless I wanted to spend hours at it, the best I could do was to get tweets sent ONLY during the last 30-45 minutes. If you used more specific search terms, or knew exactly who and what you were looking for, it would probably work more effectively but then, you wouldn’t need to search. The upshot is that search doesn’t work to revive tweets. Neither do searches on other search engines like Google, but I’ll spare you the explanations for that phenomenon.
It’s the same for Facebook, although it seems to be at least trying to help users find content, but the standard search function isn’t set up to find conversations or topics/subjects. It’s set up to find PEOPLE, because Facebook wasn’t started on the basis of focusing on content, but like many services, was based on people networking, not content sharing.
What About Facebook?
One could argue, and no doubt some will, that Twitter is just one platform for social media, and that people behave differently on it, maybe because of something specific to Twitter, like the 140 character limit. That would be a decent argument, one worth considering if it wasn’t for the fact that other platforms, Facebook, YouTube, Blogging Platforms and so on tend to show the same patterns.
Facebook is a tough one to analyse because it’s a far more complex system than Twitter. It has brand pages, and personal pages, subject pages, and walls. Facebook doesn’t treat status updates equally either. While each tweet on twitter has an equal and certain chance of appearing on the timelines of people following the Twitter account, that is NOT the case with Facebook. EdgeRank is a method Facebook uses to at least allegedly show to users what will Facebook thinks will most interest them.
It’s also the case that Facebook is less open to having third party researchers access their data, so most of the statistics about how real users use Facebook comes from them.
Still, the information that is available is not good news. Brian Carter, author of The Like Economy: How Businesses Make Money On Facebook and Facebook Marketing, and in conjunction with EdgeRank Checker, looked at 4,000 Facebook pages to determine how many people actually saw the posts.
Here’s his summarized finding:
First, we found that (in a review of 4,000 Facebook pages) the average page post is only reaching 17 percent of the page’s fans.
Brian goes on to say:
This by itself isn’t news, because there have been a number of posts across the web that talked about problems of post visibility to fans (estimating a reach per fan of anywhere from three percent to 16 percent).
But the more disturbing finding is an overall decrease in impressions per post since June 2011.
In June 2011, on average, page posts received .99 impressions per fan. That number decreased to 0.41 for December 2011. That’s a 58 percent decrease in impressions
Here’s the crux of the matter. Brian’s isn’t talking here about what is actually READ. He’s talking about the percentage of FANS (those who have chosen to read a brand’s posts) who will SEE a post. Reading, of course is a completely different issue, and while we don’t know, and probably can’t know how many status updates that appear in timelines are actually read by the people seeing them, it’s got to be very very low. Why? Because Facebook also involves a stream of updates. An update on Facebook also is time limited, and if it’s not seen, read or responded to within hours of its posting, it is also, for all intents and purposes, gone down the river.
What about “engagement” on Facebook, or to put it more specifically, what percentage of posts receive some indication that people have read them and find them valuable, by “liking” or sharing them, or getting involved in the conversation by adding a comment? Indications are that it’s nothing short of abysmal.
Michael Leander, a marketing specialist suggests that the traditionally accepted idea for a GOOD Facebook response rate is one percent, an average rate is between half a percent and one percent, while a poor rate is less than half a percent. He suggests that these accepted numbers may not be accurate, and that in fact the numbers are far lower and vary with number of followers/friends. For example, for Facebook pages with less than 10K fans/likes, the average response rate is just a tad less than one percent. At the other end of the spectrum, Facebook pages with over one million fans/likes obtain about a response rate of less than one tenth of one percent (0.09%)
One can certainly quibble with the numbers, and we should be sceptical so as not to take numbers like this as applicable across the board. Obviously response rates vary according to post, industry and so on, and these are generalizations. Regardless, the numbers are SO low, the meaning is clear. There’s much less interaction going on on Facebook than one would think. Not only is there now an issue regarding the falling number of Facebook members who SEE what is posted, never mind read them, but followers don’t seem inclined to interact with brands, despite the fact that a “like” requires almost NO effort at all.