Multi-Sector — Portfolio Holding Company
Of monthly slide decks retired across the portfolio
Reduction in KPI reporting lag after month-end
About
The company's portfolio companies wanted to replace lagging slide decks with live, anomaly-aware dashboards the board could actually trust between meetings. KPI packs arrived twelve days after month-end, teams duplicated 300+ hours of Excel effort across the portfolio, and sudden spikes were spotted only at the board meeting itself. We built a shared semantic layer feeding dual Power BI and Tableau dashboards with automatic anomaly alerts.
Industry
Multi-Sector — Portfolio Holding Company
Company size
1,000 – 5,000 employees
Founded
1988
The Company
A multi-sector holding company reporting on a twelve-day lag
The company oversees a portfolio of operating companies spanning several sectors, each producing its own monthly KPI pack for board review. Every portfolio company built its numbers independently, largely in Excel, on its own timeline.
By the time a consolidated view reached the board, roughly twelve days had passed since month-end — a gap wide enough that a spike or a developing problem in any one portfolio company could sit unnoticed until the board meeting itself surfaced it.
The challenge
Three hundred hours a month rebuilding the same numbers
Each portfolio company's finance team was independently duplicating a broadly similar KPI-pack build every month, adding up to more than 300 hours of Excel effort across the group for work that, in principle, could have been built once and reused.
Because reporting only surfaced anomalies at the point of the board meeting itself, portfolio-company leadership had no earlier warning mechanism — a sudden spike or dip was as much news to the operating company as it was to the board, discovered at the same moment rather than caught and understood in advance.
The Solution
A shared semantic layer with anomaly alerts pushed to Teams
We built a dbt semantic layer defining KPIs consistently once, feeding dual Power BI and Tableau dashboards so each portfolio company and the group centre could work from the same underlying numbers regardless of which tool they preferred to view them in. A Prophet-based anomaly detector runs continuously against the live data and pushes alerts directly into Microsoft Teams the day an issue emerges.
The shared layer eliminated the need for each portfolio company to independently rebuild its KPI calculations every month, while the anomaly alerting gave leadership visibility into developing issues well ahead of the next scheduled board meeting.
The Results
85% of decks retired, reporting lag collapsed
85% of the monthly slide decks across the portfolio were retired entirely, replaced by dashboards leadership now checks directly rather than waiting for a formatted deck. KPI reporting lag collapsed from twelve days after month-end to effectively real-time, and dashboards became a standing feature of every leadership agenda rather than a monthly artifact.
The group eliminated the more than 300 hours a month of duplicated Excel effort across portfolio companies, freeing finance teams at each operating company to spend that time on the analysis a board actually wants rather than the mechanics of assembling it.
KEEP READING
Explore more studies on AI, analytics, and enterprise intelligence.
