We’re excited to announce our new Twitter Engagement API! This API will give you the likelihood that the text in your tweet will be engaging for a standard Twitter audience.

I know what you like

Nothing is more engaging than Corgis. Nothing.

One use case of the API is to fine tune your tweets to increase the likelihood of getting retweets and favorites. Of course, the usefulness of the links and the adorableness of the pictures you share matter a ton, but this API can help you refine your writing to make your content even more engaging! Let’s say you want to promote a new blog post…

In[1] := twitter_engagement("awesome indico.io/blog/plotlines")
Out[1] := 0.40

Well, that’s worse than random chance (a 40% chance of engagement). Let’s try to add some more descriptive text.

In[1] := twitter_engagement("Blog about analyzing story structure in Disney movies 
                                  using machine learning indico.io/blog/plotlines")
Out[1] := 0.56

Great, now we’ve beaten a coin toss. Now, let’s try adding a call to action.

In[1] := twitter_engagement("retweet this blog about analyzing story structure in Disney Movies 
                                  using Machine Learning indico.io/blog/plotlines")
Out[1] := 0.61

Okay! We’re starting to really get somewhere. That’s great — now let’s add some fancy hashtags to make sure we can start trending.

In[1] := twitter_engagement("retweet this blog about analyzing story structure in #disneymovies 
                                  using #machinelearning indico.io/blog/plotlines")
Out[1] := 0.65

Now we’re getting somewhere! For a final touch, let’s throw in an @mention to see if we can get just a little extra bump.

In[1] := twitter_engagement("retweet this blog from @indicodata about analyzing story structure in #disneymovies 
                                  using #machinelearning indico.io/blog/plotlines")
Out[1] := 0.73

Sweet, now we’re really going to light up the Twitterverse!

Of course, you already know that your tweets should have descriptive text, a clear call to action, relevant hashtags, and targeted @mentions in order to be engaging. However, with this quantitative approach you can make more informed edits to help you break through the noise on Twitter. We can’t wait to see what other uses you will come up with!

Suggested Posts

Indico in the News: Deep learning takes the reins from classic machine learning

Inspiring Hack Projects: Spill.it

Make Content Recommendations Better With Machine Learning