News media companies are in the same business as Facebook and Google. They attract an audience with content and monetise that audience through advertising sales and audience revenue. In the case of Facebook and Google, the advertising revenue stream is dominant, as it was for news brands for decades. Most digital platform services are free to use — or it is better to say the cost to use those services is non-monetary. It is data received from their customers.

To optimise their businesses, platform firms such as news media companies, Facebook, and Google need to find the right value proposition and price for each of their customer groups.

Media companies should weigh several variables when figuring out where to house analytics within an organsiation.
Media companies should weigh several variables when figuring out where to house analytics within an organsiation.

For news media companies, the optimal price for each group has changed in recent years. Advertising prices have declined, and audience prices have increased. These price changes reflect the economics of advertising and digital subscription markets, as well as the value proposition news media companies are offering their customer groups.

There is an opportunity at many news media companies to optimise advertising pricing by “right sizing” the rate cards to the current market, changing the value proposition and advertising product offerings, and adding more intelligence to their B2B pricing decisions.

Audience prices have changed significantly in the past 10 years with more targeted pricing by customer segment, product innovation and bundling, and personalised subscription offers. Products have evolved as data on consumption drives resource allocation and editorial decisions in newsrooms.

A news media company that is dynamically optimising its business should work on all these areas at the same time using consistent data across the organisation. Editorial decisions should be made using data on advertising inventory demand and audience preferences. Advertising prices should depend on the available inventory and first-party data on the reachable audience on their site. Audience revenue streams should be optimised using data on propensity to buy determined by content consumption patterns.

The decision on which content should be premium or free and to whom should depend on the relative value of the audience and advertising revenue streams possible from that content to a particular segment.

So how should a news media company organise its data and analytics function? Where should this capability live within an organisation’s structure and tech stack? There are many correct answers to these questions due to each organization’s unique combination of resources, strategic objectives, and capabilities.

There are two main models: distributed “embedded” analytics staff in each core function and centralised analytics resources that provide analytics to all functions. A third approach is to create “teams” for certain products or strategic objectives that have representatives from editorial, audience, advertising, technology, and analytics.

There are benefits and costs to each of these approaches, and some companies oscillate between these models in response to challenges with their execution and changes in their strategic objectives.

Deciding on the best place for analytics resources within an organisation is a lengthy discussion, but five short pieces of advice on this topic are as follows:

  1. Put the business case first. The objective is to improve business performance. Identifying where big returns from optimisation exist within your organisation is the best first step in allocating resources and whether analytics should live close to the “front line.”
  2. Say no to shiny new things. Make sure there is a business case for each purchase that includes the total cost of ownership, including team members required to use each tool.
  3. Beware of black boxes. Many products have Artificial Intelligence features that promise to do wonderful things, but it is rare an off-the-shelf algorithm will apply well to all businesses. These features are often used to increase the cost of these applications.
  4. Don’t silo the data. Data from applications in each function should be combined with data from other functions. Combining all data in a central repository that is owned and operated by your company — in the cloud or on premises — is a best practice.
  5. Start with the end in mind. Identify the use case for analytics and work backward to understand what data is needed and where it currently lives in the organisation.

There are cases where some analytics resources are justified within a specific business function while a centralised group supports other functions with other services. A good role for centralised resources is keeping “one version of the truth” for the company, or the data set that all will use for decision making and performance reporting.

Having expensive tools purchased and supported centrally can avoid cost duplications. Analytics resources within a business function work well in cases where operational knowledge is critical to implementing analytics insights, quick response is important, and testing those insights needs to occur frequently.

Organisational design questions similar to the case discussed here with analytics also arise with questions about product design and management, financial analysis, and technical resources. There is a lot of innovation and experimentation happening in the news media business today, and organisational design is evolving quickly as well.