Delivering ERP well · Part 3 of 5
Your ERP programme doesn't have an estimating problem. It has a scope problem.
I once ran the PMO for a large bank, around 80 projects a year. The CIO believed we couldn't estimate, because projects kept coming back for more time and money. We analysed every change request. In 90% of cases, people were asking for more scope than the business case had funded.
The lesson was simple. Control scope and you control schedule and budget.
Yet on ERP, the most notorious category in IT for overruns, we buy off-the-shelf software and then customise it as if our back-office processes were unique and drove competitive advantage. They aren't, they don't, and it is very expensive.
1. Educate your steering committee first
Fit to standard only holds if the steering committee backs it. Brief them at the start. Show them exemplars of ERP programmes delivered fast and on budget, and how a fit-to-standard approach achieved it. Then show them programmes where overruns were driven by uncontrolled scope. There are plenty to choose from.
2. Inventory your old customisations
Systematically list every customisation in your current ERP and what it does. It removes the late surprise of "the old system did this, we need it adding", and lets you assess each one before deciding whether it comes across.
3. Start from standard. Customise only when it passes a test
Don't spend months designing processes before you have the platform. Take the ERP standard processes, test them against your business, and change them only where one of three tests is met:
- Competitive advantage. A change from the standard process genuinely differentiates your organisation from the competition.
- Operational necessity. Without it, the work moves into spreadsheets.
- Regulatory. The law requires it, such as e-invoicing or tax rules in specific countries.
I recently completed a programme recovery in four months. Two things made the difference: the sourcing strategy, and holding the fit-to-standard line.