Metric trees for education and edtech

Education sits in an uncomfortable space between mission-driven service and commercial reality. Universities, schools, and edtech companies all exist to improve student outcomes, yet the pressures they face pull their measurement systems in conflicting directions.

For traditional institutions, the core tension is between access and outcomes. A university that admits only the strongest applicants will have excellent graduation rates and employability scores, but it is not necessarily providing more value than one that admits a wider cohort and graduates a slightly lower percentage. Raw outcome metrics without context are misleading, which is why any education metric tree must account for the starting point of the student population, not just the endpoint.

For edtech companies, the tension is between engagement and learning. Product teams are trained to optimise for daily active users, session length, and feature adoption. These metrics keep investors and boards happy, but they do not prove that learners are acquiring knowledge or skills. A student who logs in every day but never passes an assessment is engaged but not learning. A metric tree forces these two dimensions into the same structure, making it impossible to celebrate engagement without asking what it is producing.

There is also the challenge of attribution. In healthcare, a treatment either works or it does not. In education, outcomes are shaped by factors far beyond the institution or product: socioeconomic background, parental involvement, prior attainment, motivation, peer effects. No metric tree can account for all of these, but a well-designed one acknowledges them by using contextualised benchmarks rather than raw numbers.

The root of an education metric tree should capture the dual mandate that every institution and edtech company faces: delivering genuine student outcomes while remaining financially sustainable. For a university, this might be "Sustainable delivery of excellent student outcomes." For an edtech company, it might be "Scalable improvement of learner achievement." The exact wording matters less than the structural choice to place student outcomes and organisational health as co-equal branches beneath the root.

The student outcomes branch covers academic achievement, progression and completion, and post-education outcomes. These are the metrics that define whether the organisation is fulfilling its core purpose. The institutional or business sustainability branch covers operational efficiency, financial health, and, for edtech companies, product-market fit. These metrics determine whether the organisation can continue to operate.

This two-branch structure prevents a common failure mode in education: measuring what is easy (enrolment, revenue, logins) while neglecting what matters (whether students are actually succeeding). It also prevents the opposite failure: focusing entirely on pedagogical purity while ignoring the financial and operational realities that keep the organisation running.

The student outcomes branch is where education metric trees differ most from those in other industries. Learning is not a transaction. It unfolds over months and years, and measuring it requires layering short-term proxies beneath long-term outcomes.

Retention and completion rates are the most watched metrics in education, and for good reason. They are tied directly to institutional funding, league table rankings, and regulatory compliance. In the United Kingdom, the Office for Students monitors continuation rates. In the United States, the National Center for Education Statistics tracks retention and graduation rates as a condition of federal financial aid eligibility. For edtech companies, course completion rate is the closest equivalent, and it is a critical signal of product-market fit.

But retention alone is a blunt instrument. A high retention rate could mean students are thriving, or it could mean standards are too low to fail anyone. A low retention rate could signal poor teaching, or it could reflect an institution that serves students with complex life circumstances. The metric tree addresses this by decomposing retention into its drivers.

1. 1

   Year-to-year retention rate

   The percentage of students who return for the next academic year. Decompose by programme, demographic group, and entry qualifications to identify where attrition is concentrated. Early warning systems that flag students with declining attendance or assessment submissions are the leading indicators beneath this metric.

2. 2

   Course completion rate

   The percentage of students who complete a course or module they started. For edtech products, this is often the single most important outcome metric. Decompose by course difficulty, student cohort, and time since enrolment. Low completion in the first two weeks signals an onboarding problem. Low completion in the final weeks signals a motivation or difficulty problem.

3. 3

   Time to graduation

   Measures whether students complete their programme within the expected timeframe. Longer times to graduation increase cost for both students and institutions. Decompose by full-time vs part-time status, credit transfer history, and whether students changed programme during their studies.

4. 4

   Assessment pass rates

   The most direct measure of academic achievement. Must be interpreted alongside assessment design quality. If pass rates are 98%, the assessments may be too easy. If they are 40%, the teaching or assessment design needs investigation. Decompose by assessment type (exam, coursework, practical) and by module to identify specific problem areas.

The critical insight for education metric trees is the distinction between outcome metrics and the process metrics that drive them. Retention rate is an outcome. Attendance rate, assignment submission rate, engagement with academic support services, and early assessment performance are the process metrics that predict it. A well-built tree arranges these in a hierarchy so that when retention drops, leaders can trace downward through the process metrics to identify where the problem originates.

Predictive analytics is increasingly central to this work. Many institutions now use early warning systems that combine attendance data, assessment submissions, LMS login frequency, and demographic factors into risk scores for individual students. These risk scores are themselves metrics that sit in the tree as leading indicators beneath retention, enabling intervention before a student drops out rather than after.

Edtech companies face a measurement challenge that traditional institutions do not: they must prove that their product drives learning outcomes while simultaneously demonstrating commercial viability. The metric tree is uniquely suited to this because it holds both dimensions in the same structure, connected to the same root.

The product side of an edtech metric tree borrows heavily from SaaS metrics but adapts them for the education context. Monthly active users, feature adoption, and net revenue retention are important, but they are insufficient on their own. An edtech company that grows revenue while its users show no measurable learning improvement has a business model problem that will eventually catch up with it. Procurement teams in education are increasingly demanding evidence of learning efficacy before renewing contracts.

The key is connecting engagement metrics to learning outcomes through a clear causal chain. Daily active usage matters, but only if the usage correlates with assessment performance or skill acquisition. Session duration matters, but only if longer sessions predict better outcomes. The metric tree makes these connections explicit, so that product teams optimise for the right kind of engagement, not just any engagement.

The most effective edtech metric trees include an efficacy branch alongside the product and commercial branches. This branch tracks learning gain (the difference between pre-test and post-test scores), skills mastery rates, and outcome comparisons against control groups or benchmarks. Efficacy data serves dual purposes: it guides product development by showing which features actually improve learning, and it provides the evidence that sales and customer success teams need to justify renewals and expansions.

This is not just a nice-to-have. As the edtech market matures, procurement decisions are shifting from feature comparison to evidence of impact. Institutions want to know whether a product works, not just whether it has a good user interface. The metric tree ensures that product, engineering, and commercial teams all have visibility into efficacy data, making it a shared concern rather than a research team sideproject.

In traditional industries, the connection between quality and revenue is often indirect. In education, it is increasingly explicit. For institutions, student outcomes directly drive funding, rankings, and enrolment demand. For edtech companies, learning efficacy drives renewals, expansions, and word-of-mouth growth. The metric tree makes these connections navigable.

Consider a university. Its primary revenue sources are tuition fees, government funding, and research grants. Tuition revenue is a function of enrolment volume and fee levels. Enrolment volume depends on application rates and offer acceptance rates, which in turn depend on the institution's reputation. Reputation is driven by league table rankings, which are calculated from student satisfaction scores, graduate employment rates, and research output. Every one of these financial inputs traces back through the tree to student outcomes and institutional quality.

The same logic applies to edtech companies, but through different mechanisms. An edtech company's revenue is a function of new customer acquisition and existing customer retention. Acquisition depends on marketing efficiency and sales conversion, which are heavily influenced by case studies and efficacy data. Retention depends on product engagement and, critically, whether the product is delivering measurable learning results. When an institution renews an edtech contract, the question is increasingly "Did student outcomes improve?" not "Did teachers like the interface?"

This second tree shows the edtech commercial model decomposed. Notice that learning outcome metrics appear on both sides: efficacy case studies drive acquisition, and student outcome achievement drives retention. This is the structural argument for investing in efficacy measurement. It is not a cost centre. It is a growth driver that feeds both sides of the revenue equation.

The metric tree also reveals a common failure mode in edtech. Companies that optimise acquisition without investing in efficacy will grow quickly but face a churn problem as customers discover the product does not deliver on its promises. Companies that invest heavily in efficacy but neglect go-to-market will have great outcomes data but insufficient revenue to survive. The tree holds both in view, ensuring neither is neglected.

Between student outcomes and financial sustainability sits a layer of operational metrics that determine how efficiently an institution converts resources into educational value. These metrics are often overlooked in favour of either the headline outcomes or the bottom-line finances, but they are the levers that leaders can most directly influence.

Student-to-faculty ratio is one of the most scrutinised operational metrics in education. It appears in league tables, influences student choice, and directly affects teaching quality. But it is also a cost driver. Lowering the ratio improves outcomes but increases salary expenditure. The metric tree makes this trade-off visible by showing the ratio connected to both the outcomes branch (through teaching quality proxies) and the financial branch (through cost per student).

Facility utilisation is another metric that benefits from tree-based thinking. Lecture theatres, laboratories, and study spaces represent significant capital investment. Low utilisation means these assets are underused, increasing cost per student hour. But high utilisation can mean overcrowding, which harms the student experience. The optimal point depends on the type of space: a lecture theatre at 95% capacity is efficient; a laboratory at 95% capacity is unsafe.

Building a metric tree for an education organisation follows the same fundamental principles as any other metric tree, but the specific considerations differ depending on whether you are an institution or an edtech company. Here is how to approach each step.

1. 1

   Define the root around student success, not revenue

   For institutions, the root should reflect the educational mission: "Sustainable delivery of excellent student outcomes" or "Equitable student success and institutional viability." For edtech companies, orient the root around learner achievement: "Scalable improvement of learner outcomes." This ensures the tree structure subordinates commercial metrics to the purpose they serve.

2. 2

   Split into outcomes and sustainability

   The first decomposition should separate student outcomes (achievement, retention, post-education success) from organisational sustainability (efficiency, finance, and for edtech, product health). This prevents the common failure of treating financial metrics as the only tree and bolting student outcomes on as an afterthought.

3. 3

   Layer leading and lagging indicators

   Graduation rates and employment outcomes are lagging indicators that take years to materialise. Attendance rates, assignment submission rates, and early assessment scores are leading indicators that predict them. Place lagging indicators higher in the tree and leading indicators beneath them, creating a clear path from early signal to eventual outcome.

4. 4

   Contextualise with benchmarks, not just targets

   Education metrics are meaningless without context. A 70% retention rate might be excellent for an open-access institution and poor for a selective one. Include contextual benchmarks (peer group averages, value-added measures, prior attainment baselines) alongside absolute targets for every metric in the tree.

5. 5

   Assign ownership across academic and administrative lines

   Academic metrics need academic owners: heads of department, programme directors, student success teams. Operational metrics need administrative owners: finance directors, facilities managers, IT leaders. Where metrics span both domains, such as student-to-faculty ratio, establish shared ownership with clear accountability for each dimension.

6. 6

   Start with one programme or product line

   A university-wide metric tree covering every department, programme, and support function is overwhelming. Start with a single programme or faculty. For edtech companies, start with your core product. Prove the value of the connected model in a contained scope before expanding. A metric tree that people actually use for one programme is more valuable than a comprehensive diagram that no one consults.

The most common mistake in education metric trees is including too many metrics. Institutions and edtech companies track hundreds of data points. The tree should contain 20-30 that matter most for strategic decisions. Everything else belongs in departmental dashboards or product analytics tools as supporting detail.

A second common mistake is treating the tree as static. Student populations change. Curricula evolve. Product features are added and removed. The metric tree should be reviewed at least annually, and the connections between metrics should be validated with data. If your tree says that attendance predicts retention but the correlation has weakened, the tree needs updating. A metric tree is a model of how your organisation works. Like all models, it must be tested and refined.