Large-scale application and technical upgrades: three technology leaps!

Client: Securitas — Services sector.
Three successive technology leaps, in Big Bang mode, across two business applications (HR/Payroll and Finance), to clear technical debt without the cost of a re-implementation.
Context
- Heavy technical and functional debt requiring a complex application and technical upgrade.
- Finding an alternative to the upgrade cost estimated by the publisher and the IT services firm, which matched an initial implementation.
- Version upgrade of the HR/Global Payroll and Finance applications, meeting high-availability constraints.
- Limiting production downtime to a single weekend.
- Keeping the HR and Finance databases in sync to avoid building transitional interfaces.
Levers
- Finance application: 2 leaps (V8 → V9 → V9.2) of the GL, AP, AR, AM, BI, PC, CA modules.
- HR/Payroll application: 2 leaps (8.3 → V9 → V9.2) of the GA, Training, Global Payroll modules.
- Technical leaps: 8.20 → 8.49 → 8.54.
- Rebuilding the technical repository to obtain a project/component mapping.
- Review and migration of the configuration repository.
- Technical analysis reports for fit/gap workshops and for carrying over code and its impacts on interfaces and reporting.
Approach
- A methodology involving three technology leaps in a Big Bang project spanning two business applications.
- Command of the right-sized data estate.
- Estimation of development-carryover effort — optimised carryover.
- Building the technical and functional repositories.
- Technical, functional and data-model audits of acceptance environments at each Move to Prod.
Results
We went from 2,500 days to 1,400 days — a substantial 40% saving. The project close-out is a success: an original approach, tangible economic superiority, a 300% optimisation ratio and an aligned business-IT pairing.
— Christophe Guéguen, CIO, Securitas
- Three mastered upgrade waves: 2009 (8.0 → 8.3 → 9.0), 2012 (9.0 → 9.2, −40% on total project cost) and 2016 (PUM & Tools 8.55, −50% on total cost).
- Optimised lifecycle: 17-month project in 2009, 9 months in 2014.
- Overall 40% saving (from 2,500 to 1,400 days), a 300% optimisation ratio.