ACM Data Team
Overview of this campaign
In mid 2018 Australian Community Media (ACM) launched a digital subscription product on 42 of their regional news websites. In late 2018 the ACM data team was founded to create a data warehouse and perform BI reporting for the new subscriptions business unit and other digital revenue streams.
The data warehouse was built as a greenfield project, utilising low-cost cloud technology from Google Cloud. A diverse range of data sources were connected to via API, information on users’ subscriptions, payments, interactions on the website, email campaign responses etc.
This data was transformed in the data warehouse and exposed to the business via business intelligence dashboards with Google Data Studio. Dashboards are provided for the following business streams:
Management - financial/growth reporting for subscriptions
Editorial - user engagement, article performance
Marketing - marketing channel effectiveness and ROI
Product - A/B test results
Sales - advertising viewability, sell-through rates
Key objectives of the data analytics team were to provide real-time dashboards to editorial, provide business analysis on pricing/discount options and to monitor the impact of subscriptions on digital advertising revenue.
Once a single source of truth was created and all data sources integrated then predictive analytics was used. Subscriber and revenue growth is forecasted using regression analysis. Subscriber propensity to churn is forecasted by separating users into profiles based on metadata about their subscription (subscription type, payment method, device used etc). News articles are categorised using natural language processing to compare conversion volume across different content types.
Integration of data into business processes and culture was a key consideration. Use of standardised reporting allowed greater transparency and accountability when assessing the impact of digital initiatives. Weekly data meetings are held with all stakeholders to assess the progress of subscription north star metrics such as ARR and subscriber growth.
Results for this campaign
The creation of a single source of truth and dedicated business analysis support for subscriptions allowed the design of a north star metric for the new subscriptions business. At launch Annual Recurring Revenue was chosen as the north star for subscriptions, taking the price of subscriptions, quantity of subscriptions and user retention into account. This was chosen because it identifies three levers the business can act on to deliver revenue growth.
Clarity around reporting was achieved by widely publishing the methodology for other calculated metrics such as churn and subscription growth. All subscription metrics were standardised to 30 day rolling windows meaning that for any masthead it is possible to compare diverse metrics such as churn, conversion rate, % of engaged users etc on a like-for-like basis. This unifies data reporting for 42 subscription sites and 100+ free sites and reduces confusion around competing methodologies and time frames.
Promoting data analysis as an in-house capability allows the business to not be reliant on external consultants for analytical services, this realised significant cost savings and provides future opportunities for the business. The data warehouse is run on very low infrastructure cost (currently $5,000 a year) due to efficient use of cloud resources and free resources.
Providing data reporting in-house also allowed far greater personalisation for internal stakeholders; journalists and editors have access to personalised dashboards. This is enabled through the organisation’s use of Google, all staff accounts are Gmail accounts and access to dashboards and databases is controlled via Google authentication. Data team dashboards are used by approximately 400 individuals across the company with user groups in sales, marketing, editorial, product and management.
The efficiency of testing changes to digital product has been increased as the methodology for A/B testing site changes, subscription prices and marketing messages is automated through the warehouse. Testing can be automated for specific user segments created in the warehouse, examples include variations of cart abandon emails and SMS messages automatically sent to users filling set criteria.