Financial Variance Analysis for an Online Education Company

The situation

An online education company with 18 departments had business-critical data spread across multiple systems, and no integrated view of how the business was performing against plan. Financial projections lived in spreadsheets maintained by the finance team. Enrollment data lived in a student information system. Revenue data lived in the CRM and billing systems. When leadership asked whether the company was on track, the answer required days of manual data gathering and reconciliation.

The company was growing rapidly and preparing to launch a new academic program that would add significant complexity to financial planning. Leadership needed to understand not just whether numbers were on track, but why they were or were not. A revenue miss could stem from enrollment shortfalls, pricing changes, retention issues, or timing differences. Without a system that connected financial projections to the operational assumptions underneath them, every variance investigation was a manual forensic exercise.

What we built

CorrDyn built a data engineering layer that integrated data from the company’s CRM, billing system, student information system, and operational databases into a governed warehouse. We used dbt to define transformation logic and Fivetran to maintain automated data pipelines from source systems. Census handled reverse ETL, syncing calculated fields and segment assignments back into HubSpot for the marketing and enrollment teams.

On top of the integrated data, we built a Tableau reporting suite organized around the relationship between financial projections and operational reality. The core dashboard tied each line item in the financial plan to the operational assumptions that produced it. Enrollment projections were decomposed into lead volume, conversion rates, and start dates. Revenue projections were decomposed into enrollment counts, pricing tiers, and retention rates. Each assumption had a corresponding dashboard view showing actual versus plan with trend lines.

A series of supporting dashboards tracked the operational metrics that fed the financial model: enrollment pipeline by program and cohort, retention and completion rates, marketing channel performance, and departmental KPIs. These dashboards were designed so that when the financial dashboard showed a variance, leadership could drill into the operational dashboard to identify the root cause without leaving the Tableau environment.

We also automated several processes that had been consuming hundreds of hours per month across departments: scholarship funding calculations, course roster generation, invoice creation, and revenue reconciliation.

What changed

Leadership shifted from quarterly review cycles to daily performance monitoring. When the financial dashboard showed enrollment revenue tracking below plan, the team could immediately see whether the issue was lead volume (marketing), conversion rate (admissions), or retention (student success). This specificity turned variance discussions from blame-assignment exercises into diagnostic conversations about which assumptions needed updating and which interventions would have the most impact.

The process automations freed staff time that had been spent on manual data manipulation. The company grew revenue 87% in a single year without corresponding increases in cost or headcount. Finance leadership described the forecasting process as “vastly more productive,” spending time on strategic analysis rather than spreadsheet preparation. The engagement has continued for over nine years, expanding to cover new programs, new data sources, and new analytical capabilities as the company’s needs evolve.