Businesses and individuals alike fret over what to say on Twitter, whether it’s too offensive, or too much like a marketing pitch. But according to a new algorithm, there’s actually a way to figure out the “perfect tweet”.
And according to a new piece of research, this may well be it:
At least, according to one new paper from UCLA and HP Labs. The two have developed an algorithm to find what makes up the perfect tweet – and they’ve actually managed to score a very high accuracy mark of 84%.
To develop the algorithm, researchers created a hypothesis that four factors would determine the success of a tweet, including the news source of the tweet, the category of news, whether the language was emotional or objective, and whether celebrities or famous brands were mentioned.
They then looked at 40,000 news article tweets, and then assigned a category to each article. Then they tried to recognise names in the tweets, and then scored each individual message.
The researchers found their hypothesis was correct, so a tweet about someone famous will mean something more coming from a reputable source than an individual. That much is obvious, but putting all four factors of the hypothesis together comes up with an algorithm for success.
As The Atlantic points out, behaviour around brands works the same on Twitter as it always has.
“What’s remarkable about the HP paper and its algorithm is how back-to-the-future their results actually are.”
“Online, the researchers are saying, the power of the brand is exactly what it has been since brands first emerged in the Middle Ages: It’s a vector of trust. And it’s a direct proxy for the ongoing transactional realities of the in-person human relationship. “
If you’re interested in how and what works in the Twitter universe, then this is well worth a read.