Blog

Building a business case for learning analytics: securing stakeholder engagement and ongoing support

James Hodgkin
by
James Hodgkin

I’m concluding my two-part blog series reviewing what you need in the business case for your project, to make it an established part of your organisation’s business-as usual.

A female tutor helping a male student with a laptop.

In part one of my learning analytics blog series, I considered the implications as well as the launch of a project, including the foundations, governance, technical readiness and operability. I’m concluding part two with a deeper diver into the people behind the protocols.

Optimism bias is a real risk to project success: it’s essential to forecast realistic return on investment (ROI) and the cost of existing resources and any new requirements, which means including the people behind the process.

Phasing and delivery

Stakeholders and teams are crucial to the success of your business case – from approval through to business-as-usual. To ensure a successful implementation you must include the impact on already busy teams:

  • Start with a pilot in one faculty or a defined cohort such as first year undergraduates
  • Gather staff and student feedback, refine thresholds and workflows, and scale in phases
  • Design role-based training for tutors, student support teams, registry and compliance officers
  • Communicate with students transparently about purpose, benefits, data use and support - build digital confidence, not anxiety
  • Liaise with your planning team to ensure the alignment of management information with learning analytics

Options appraisal, benefits, ROI and the cost of success

Quantify outcomes and plan for the implications of doing this well:

  • Track continuation uplift, avoided withdrawal costs and compliance timeliness
  • Options appraisal may include a breakdown of the projected cost of building in-house rather than buying an ‘off-the-shelf’ product
  • Bring together the right institutional data to measure student outcomes after interventions
  • Allocate time to analyse how interventions impact engagement
  • Measure processing times - a single view reduces time spent navigating systems and increases time with students
  • Anticipate increased identification of need. Use prioritisation and notification patterns that direct effort where it matters most

Illustrative calculation for your finance paper:

  • If continuation improves by 1.0 to 1.5 percentage points in targeted cohorts and the average lost income per withdrawal across a three-year degree is up to £27,750, avoided cost equals the number of students multiplied by the percentage point uplift multiplied by the lost income estimate. A simpler way to say this could be if early intervention saves ‘x’ number of students, you’ve saved ‘£x’, but base your framing on the audience you want to reach
  • Add time savings - for example, 30 to 45 minutes per week per tutor multiplied by the number of tutors and a salary proxy. Support staff time is more likely to reduce by hours once the system is an established part of business-as-usual
  • Offset against total cost of ownership for licensing, implementation, training and ongoing support

Risks, ethics and safeguards

Build trust:

  • Involve student union and subject reps in design and communication: open, transparent information is essential to maintain trust and understanding of the data being collected and what it is used for.
  • Explain indicators and provide recourse if a student challenges an alert: qualified staff should understand, validate, review, and improve all algorithms and metrics used for predictive analytics or interventions.
  • Review models regularly for bias and false positives. All metrics and algorithms must be peer reviewed and validated on a regular basis
  • Set notification standards to avoid over alerting and focus on proportionate outreach

The approval essentials

Make it easy for committees and programme boards to approve:

  • Executive summary with purpose, scope, outcomes, costs and timeline
  • Implementation plan with phases, milestones, resources and dependencies
  • Policy impacts across tutoring, engagement, attendance and privacy notices
  • Financials with total cost, benefits, ROI and sensitivity analysis
  • Risk and mitigation with ethics, data, capacity and integration controls
  • Evaluation framework with key performance indicators (KPIs), reporting cadence and governance owners

Find out more

About the author

James Hodgkin
James Hodgkin
Head of analytics, Jisc