How To Conduct Headline Variations?

The process for conducting headline testing is to set up the original headline as a control, and choose at least one or two variants. Viral media sites such as Upworthy test as many as 25 headlines per article, but many sites do not have large enough audiences to justify testing that many variants, since it would take a significant amount of traffic to get a statistically significant sample size.

In order to display the headline variants simultaneously to website visitors, an A/B testing tool is used to split the traffic that reaches a website and display the different variations of the headline. The A/B testing platform will gather data from the test, and declare a winner once the a sufficient number of people have viewed the test for the results to be statistically significant.

Once the headline test has determined a winner, 100% of the traffic can then be directed at the winning variation. Since the headline of an article is one of the key things that readers look at when reading or sharing an article, optimizing the headline can result in significant boosts in readership and other metrics.