New Data: “Engage in the Conversation” May Not Actually Work

For more mythbusting and social media science, don’t forget to register for my Science of Social Media webinar.

If you’ve seen me speaking, or you read my blog regularly, you may know that one of my favorite past times is busting social media marketing myths, and one of my favorite examples of “unicorns and rainbows” advice is “engage in the conversation.” Finally, I decided to use data to test the notion that “engaging in the conversation” on Twitter has a positive impact.

People follow you on Twitter because they like what you’re saying, and the more people who follow you, the larger your reach is. If you are a marketer, there is no point in pursuing a twitter strategy that doesn’t include getting as many followers as possible. Naturally, one of the most important measures of success on Twitter is follower count, and I compared that to the replies percentage–the percentage of your Tweets that start with an @ sign–of a random selection of over 130,000 Twitter users.

What I found may surprise you. When you look at the average following of users over their reply percentage, you see that users at the extremes–those who reply rarely or constantly–have far more followers than those users in the middle of the graph–users for whom replies make up 30%-50% of their tweets.

More tellingly however is what you find when you look at the average reply percentage of folks with over 1,000 followers and compare it to the reply percentage of users with less than 1,000 followers. Users with lots of followers respond much less frequently. The effect is the same when you compare users with more than 1,000,000 followers with those that have less.

I want to be careful about causation here, as users with many followers may respond less simply because they have too many mentions to reply to. But I think this data does start to question the knee-jerk unicorns and rainbows myth about “engaging in the conversation.” What do you think?