Two institutions can ask for the same thing, an education analytics dashboard, and receive quotes that seem to describe different projects. The difference is rarely the chart library. It is the number of source systems, the state of the student and assessment data, the roles that need access, the refresh frequency and how much validation the institution expects before a number is trusted.
This guide breaks the cost into components, gives a percentage model for planning and provides a calculator-style formula you can fill with your own inputs. It does not quote a market price, because a defensible budget depends on scope, not on a rate card copied from another country.
Why quotes diverge
Behind one enrolment chart sit a student information system, attendance records, assessment marks, fee records and perhaps a learning management system. Each source has its own identifiers, definitions and quality problems, so a quote that assumes clean data is not comparable with one that includes integration and reconciliation.
Ask every vendor to state what they assume about the sources. The assumptions are the quote.
The seven cost lines
1. Discovery and metric definition
Workshops with academic, finance and operations staff to define enrolment, retention, attendance, pass rate, collection and workload metrics. The output is a metric dictionary with owners, formulas and filters. Skipping this line produces a dashboard that looks right and is argued about in every meeting.
2. Data integration
Extraction from each source, mapping of student and programme identifiers, handling of historical records, time zones and academic calendars. Integration effort rises with the number of sources, the age of the systems and the absence of stable keys.
3. Data model and warehouse
A store that combines sources at a consistent grain: student, term, course, section, assessment and transaction. The model needs history so that a change in a programme or fee structure does not rewrite past reports. It also needs documentation that survives staff turnover.
4. Dashboard build
Pages for different audiences: leadership, academic coordinators, finance and admissions. Cost depends on the number of distinct views, the interaction required, mobile use and whether the institution wants narrative commentary alongside charts. A single executive page is a different project from a self-service exploration layer.
5. Identity, roles and privacy controls
Single sign-on, role definitions, row-level restrictions and audit logs. A teacher should see their sections; a coordinator should see their programme; leadership should see aggregates. Student-level data raises privacy obligations, so access design is not an afterthought. The Bangladesh data protection context and internal policy both shape what can be shown and to whom.
6. Hosting, licences and tooling
Whether the dashboard runs on a BI platform, a custom application or a module inside a coaching-centre system. Licence models vary by user, by server or by consumption. Hosting may be local or cloud, with a cost for backup, monitoring and environments.
7. Validation, training and support
Reconciliation against source reports, user acceptance testing, training for each role, documentation and a support arrangement after launch. This line is often underestimated, then funded from the next year's budget under a different name.
A percentage model for the build
Use these shares to sanity-check a proposal, not to set a price. They assume a custom dashboard over existing systems, with moderate history and multiple roles. Shift a few points between lines as scope changes, but keep the total at 100.
- Discovery and metric definition: 10% of the one-off budget.
- Data integration and modelling: 35%.
- Dashboard and interaction build: 20%.
- Security, roles and audit: 15%.
- Validation and training: 10%.
- Contingency: 10%, held by the institution rather than the vendor.
If a quote puts less than a quarter into data and security, ask what it assumes. If it puts almost everything there, ask what phase one actually delivers.
A calculator for the full envelope
Budget the first year and the annual run rate separately. A simple formula:
Year-one cost = build one-off + hosting and licences + support + internal staff time + contingency
Annual run cost = hosting and licences + support + internal analyst time + annual review and training
Fill each input with your own scope:
- Build one-off: the sum of the seven lines above, based on your source count and dashboard count.
- Hosting and licences: the vendor quote for your expected number of users and data volume.
- Support: a retained hours block or a service-level agreement, priced separately from new work.
- Internal time: hours from academic and IT staff for workshops, validation and training, multiplied by a loaded internal cost.
- Contingency: a share of the one-off build for source surprises and scope correction.
Do not compare vendors on the build line alone. A low build with a high licence and support load can cost more over three years than a higher build with a simple run model.
Scoping inputs that move the number most
- Source systems: each additional system adds mapping, testing and reconciliation.
- History: five years of archived records may need cleaning before they can be compared.
- Refresh frequency: nightly batch is simpler than near-real-time; intraday refresh adds engineering and monitoring.
- User roles: each distinct permission model needs design and testing.
- Student-level access: row-level security and audit requirements increase build and review effort.
How to defend the budget
Present the budget as a decision with options: a minimum viable dashboard, a standard build and an extended build. For each, list the questions it answers, the sources it uses, the roles it serves and the annual run cost. Tie the spend to a recurring decision such as admissions planning, section allocation or academic intervention. A dashboard that changes no decision is a cost without a return.
Questions for a vendor
- Which source systems have you integrated before, and what broke?
- Who owns the metric definitions, and how are changes approved?
- How will you prove the numbers reconcile with the source reports?
- What is included in support, and what is quoted as new work?
- How do you handle student-level access and audit logs?
- What will the institution own at the end, and in what format?
A 90-day path to a usable number
- Days 1-15: list sources, owners and the five decisions the dashboard must support.
- Days 16-30: define metrics and produce a sample reconciliation by hand.
- Days 31-50: get two vendor quotes against the same scope and assumptions.
- Days 51-70: build a fixed-scope prototype for one audience and one source set.
- Days 71-90: validate with users, price the annual run, and decide whether to proceed or stop.
Metrics to hold the spend accountable
- Share of target users active each month.
- Data freshness against the agreed refresh window.
- Reconciliation variance between dashboard and source reports.
Scoping help for education teams
If you need help scoping a dashboard before you commit to a build, start with Education Operations. The pricing page explains how discovery, implementation and support are separated. For the wider measurement design, read the KPI dashboard buying guide.
Sources
- DataGuidance, Bangladesh data protection overview, accessed 17 November 2026.
- UNESCO Institute for Statistics, accessed 17 November 2026.
- NIST Privacy Framework, accessed 17 November 2026.
- Metabase documentation, accessed 17 November 2026.
- PostgreSQL documentation, Materialized views, accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.