Q: Why has FT Strategies launched a Data and Insights Practice?
A: I can’t think of a single publishing organisation that doesn’t have challenges when it comes to having reliable and useful audience, financial or industry data. Whether it is implementing web analytics, managing content meta-data or having a single customer view, there are always improvements to be made.
The Data & Insights Practice’s goal is to help clients — both at a strategic and very practical level — solve these problems. To do that, we have hired some of the best data scientists, analytics experts and researchers; we hope this investment will increase the value that clients get from our projects.
Q: What recent innovations in data science are most relevant to publishers?
A: There are too many to mention, but here are a few exciting ones:
- Personalised homepages: At a recent event, Aftenposten talked about how their data science team built a homepage algorithm that improved CTR by 25%. The customer algorithm factored in: news value (self-rated by individual journalists), subscription conversion rates, audience engagement and personal interest (of the audience member). What makes this innovation so interesting is not only the results but also the fact that it includes self-declared, audience and personal data.
- Dynamic access models: Although there are some really exciting examples of success, the amount of effort to implement these models is still very high. We are continuing to work with our clients and FT to understand how to implement a dynamic paywall and whether that is the right decision for the organisation.
- User needs tagging: The concept of user needs has been around for a long time, however LLM-driven recommendations are pretty new. That is what Smartocto (who we have recently agreed a commercial partnership with) are doing — providing analytics tools that include personalised recommendations to help newsrooms better meet user needs. It’s exciting to see how this can have a downstream impact on overall user satisfaction and engagement.
Q: Where do publishers most often struggle with data strategy?
A: The biggest challenge that publishers face when it comes to their data strategy is evidencing ROI (return on investment). Given how resource-constrained many news organisations are, they are often reluctant to throw money at expensive SaaS without knowing what they will get in exchange. Without money or senior management buy-in, data strategies (no matter how good) often fail to go beyond documents or presentations.
From our experience, the publishers with the strongest data capabilities can estimate the value of having high-quality data assets. For example, at the Financial Times, we know that building direct relationships with our users and creating a login (and single user ID) generates meaningful revenue for the business and that if we achieve our ‘recognised reader’ goal, that will translate to millions of pounds in subscription and advertising revenue. These kinds of data points allow us to invest in the technology and skills that underpin an effective data strategy.
Q: What are the most overlooked data signals that can indicate the health of a business?
A: Although it’s rarely overlooked now, engagement is by far and away the most important health indicator within a publishing business. From our experience, engagement is the perfect ‘leading’ data point in that it is predictive of acquisition, churn and customers taking desired actions (eg. upsell).
A measure that is often talked about, but less often calculated, is customer lifetime value — ie. the predicted amount of revenue that an individual will contribute to the business based on expected churn, price and ad exposure. When calculated at a user segment level, we often find clients are extremely surprised by the difference between groups - eg. a registered user is often 10x more valuable than an anonymous user by virtue of their propensity to subscribe and ad exposure. We regularly help clients calculate this for the first time as it is a very important health indicator.
Q: Beyond basic analytics, what advanced metrics should publishers track?
A: My favourite, which very few publishers know about is ‘normalised churn’. Normalised churn, rather than using the total number of subscribers as the denominator, uses the total number of renewals that were due in the timeframe as a baseline. This allows subscription businesses to differentiate between whether any changes in their churn rate are the result of limiting the number of renewals or improving the actual renewal rate. You must distinguish between those two factors and measure them as they are often the responsibility of different parts of the organisation. Often publishers are much better off reducing churn at the point of acquisition by incentivising subscribers onto annual products (8-10x higher LTV) than adopting improved churn strategies.
Q: Can you share an example of how a data-driven approach has led to a significant business transformation?
A: Last year, we ran a data-driven project with a leading European news organisation to help them understand and reduce subscriber churn. As part of that project, we helped to build a custom ‘propensity to churn’ data science model to inform their pricing and renewal strategy. Since working together, that organisation has reduced churn by 50% - leading to a meaningful growth in digital subscribers to over 500,000 this year.
This case study, however, does highlight an important point: the impact of data-driven initiatives are only powerful if they are used to inform and shape business decisions or initiatives.
The Data & Insights Practice at FT Strategies is a specialist group that offers data implementation services, bespoke research, data analytics, business & audience benchmarking and custom thought leadership.
Website: www.ftstrategies.com
