ML Analytics Raises Higher Education Division Revenue by £100M During COVID

Context

A £1bn global higher education corporation required a strategic response to the disruption caused by the Covid pandemic (2019–2021). The organisation operated across multiple countries with more than 10,000 staff and a diverse portfolio of teaching and operational models. Leadership needed a single, reliable source of insight to guide student-facing strategy, operational planning and revenue management during an unprecedented period. The mandate was to combine internal operational data with carefully curated external indicators, apply machine learning and predictive analytics, and deliver a governed, compliant analytics capability that would steer decisions through the pandemic and beyond.

Challenges

Three core challenges defined the programme. First, data quality and organisation: data resided in numerous departmental silos with inconsistent definitions and variable quality. Second, the use of machine learning and predictive analytics on external data introduced additional risk — not all external data sources were usable, and combining them with internal records required careful vetting and harmonisation. Third, scalability and complexity across a large, global organisation with more than 10,000 staff: the solution had to operate across different regulatory regimes, languages and operational practices, and handle peak loads without disruption. Country-to-country differences (regulation, enrolment patterns, local operations) created multiple layers of complexity that needed to be engineered into the analytics design from the outset.

Implementation

The project consolidated siloed data into a single, governed and compliant data source that became the authoritative foundation for analysis. A central data governance framework standardised definitions, lineage and access controls to ensure regulatory compliance across jurisdictions. The analytics team integrated internal operational systems with a tightly curated set of external data sources; each external feed was evaluated for reliability and usefulness since many potential sources were discarded as unusable. Machine learning models and predictive analytics were applied at a high strategic level: forecasting demand, modelling cost streams, and simulating scenarios to guide both revenue strategy and internal workforce planning.

Country differences were explicitly modelled — regulatory constraints, local operational practices and regional risk profiles were parameterised so outputs respected local realities while remaining comparable at a corporate level. Scalability was pre-planned: the programme worked directly with vendors to guarantee capacity 10X beyond initial requirements, ensuring that the platform could absorb large-scale experiments, increased data volumes and future expansion. This vendor collaboration and capacity planning future-proofed the solution for decades, allowing the organisation to adapt quickly as circumstances evolved. The resulting capability was positioned as a strategic analytics tool rather than a narrow operational dashboard, empowering senior leaders with forward-looking, probabilistic insight during a period of high uncertainty.

Results

The strategic insights derived from the machine learning tool contributed to a significant commercial outcome: division revenues grew from £650m to £750m during the pandemic, an outstanding performance given the sector-wide disruption. Predictive cost-stream modelling improved workforce planning, enabling the organisation to reduce overall workforce costs through targeted efficiencies without compromising service delivery. Importantly, students continued to receive an excellent education service throughout the crisis; operational changes informed by the analytics capability helped protect teaching quality and student experience while adapting to public-health and regulatory constraints.

Beyond immediate gains, the project delivered longer-term value: a governed, compliant single source of truth that supports decision-making across countries; scalable architecture designed to grow well beyond initial needs; and a proven methodology for assessing and integrating external data. Together these outcomes positioned the organisation to respond nimbly to future shocks, optimise workforce and cost structures, and sustain revenue growth while maintaining high standards of education and student support.