8 min read

Why we developed ROSA

Ryan Bradley

If you’re a content creator or work in publishing then the Google News Initiative Data Lab APAC 2020 Playbook is a must read.

The playbook was borne out of the GNI APAC Data Lab, a publisher-centric program that aims to help Asia-Pacific publishers transform from “developing” to “mature” by adopting data-focused processes.

At Incites we’re especially pleased that our ROSA framework was cited as a best practice by Google for content planning.

What is the ROSA framework and what does it mean for your business?

How does Incites determine which content to Replicate, Optimize, Stop or Amplify?

The ROSA framework is a data-driven approach to developing actionable insights. Not only does our ROSA framework provide insight into what is currently happening within organisations and markets, it also provides actionable recommendations to further enhance workflows and reduce operational costs.

We came up with this framework by defining the overarching business objective (e.g. a specific conversion outcome), generating a score for each article in relation to this business objective, and comparing this with scored metric(s) which represent how well content is reaching their audience. At the core of our ROSA framework is our scoring algorithms which score data points on multiple criteria defined by business objectives. Based on these scores, data points can then be plotted on a plane and use its associated location to drive the actionable ROSA insights.

For example, the “Replicate” articles are identified on the basis of well they outperform other articles in terms of driving brand-lovers, registered or subscribed users both in terms of audience size and average time-spent.

“Optimize” articles over-index in terms of attracting an audience but under-perform in terms of driving outcomes like engagement and recirculation.

“Stop” articles are those that attract little to no audience and do not engage users. Establishing a pattern in terms of common characteristics of these articles is key to ensuring that resources are not wasted on these types of stories.

“Amplify” articles over-perform in terms of engagement and recirculation but have not over-performed in terms of traffic or audience size. Consider what would be the most effective channels to re-promote or redistribute this article.

With, we use ROSA to literally tell storytellers what stories to tell

The platform helps publishers grow reader revenue by delivering actionable AI-powered insights, written in simple natural language, to their newsrooms.

Instead of dashboards, the ROSA insights are delivered as actionable messages directly into the editorial workflow eg. as messages posted automatically into Slack, MS Teams, Content Management Systems and anywhere work happens.

Written by
Ryan Bradley
Data and Software Engineer

Ryan utilises his approx. 7 years of experience at the intersection of computer science, enterprise-grade software engineering, and data science to gain an understanding of the world from many knowledge-domains and industries. He not only builds innovative scalable insights based on consumer and user behaviour, but also manages, educates and consults teams and business as a whole on data, best practices, and how these behaviours can have an effect on the user/individual, business objectives, growth and culture.