Risk Models to Proactively Manage USD 2.5B Annual Construction Project Portfolio

Business Objective

Our client is a leading US-based construction company specializing in highly complex and sustainable projects such as data centers, tech offices, and green buildings.

Their objective was to predict the success of their new and ongoing projects along two dimensions – Budget and Time. A project is a Budget(Profit) success if it achieves or exceeds estimated profit. Likewise, a project is a Time success if it achieves or beats contract completion date.

  • Need to predict project success:
    • Accurately without being too alarmist or too lax – in other words, high accuracy and lower false positives
    • Early in the life cycle to allow time for corrective action
  • More than 200 project and financial attributes to consider
  • Missing data, errors, and inconsistencies in manual entry of data by project managers.

Solution Methodology
  • Explored different risk attributes from the data, their distribution, and impact on project success – this included historical data from monthly status reports of closed projects with known outcomes
  • Combined several attributes and provided drill-downs to identify types of projects which were at highest risk
  • Captured key metrics– such as project budget overrun – and how they correlate with risk attributes
  • Built a predictive model which quantified the impact of shortlisted factors on project success.
  • Built an early warning dashboard in Tableau that identifies risky projects in a portfolio for business action. The system can:
    • Provide alerts from the first month of the inception of a project
    • Update predictions every month on a rolling basis.
Business Impact
  • Through the early warning system, the client was able to proactively risk manage USD 2.5B per annum worth of construction portfolio.

Copyright © 2023 Tiger Analytics | All Rights Reserved