Shipping ERPs that people actually use
Most ERP projects fail at adoption, not at launch. Three patterns from our deployments that keep usage high after go-live.
An ERP nobody uses is worse than no ERP — you've paid for the system and kept the spreadsheets. Across our deployments, adoption is the metric that decides whether a project succeeded.
Pattern one: build the workflow people already have, not the workflow the org chart implies. We shadow real users in discovery and encode their actual process — then improve it where it's safe to, not everywhere it's possible to.
Pattern two: role-shaped interfaces. A warehouse operator and a finance director need different screens of the same system. When the UI matches the job, training time collapses.
Pattern three: go-live is the midpoint, not the finish. The first month in production surfaces the real edge cases. We budget support into every deployment so fixes land in days, not in a 'phase two' that never comes.
The pattern underneath all three: treat the ERP as a product for its users, not a deliverable for its buyer. That's the difference between software that runs a company and software that runs out of budget.