"The most expensive software I ever bought was the one nobody opened." That's how an auto-parts owner in Faridabad describes his first ERP: a big-brand system, a big invoice, and within six months a floor that had drifted back to its registers and spreadsheets. The second attempt worked. Here's what changed.
Mistake 1: Buying for features you'll never use
The first system could do everything, which meant it did nothing simply. SMEs don't need a hundred modules. They need the ten screens their team will touch every day to be fast and obvious. He chose the second ERP on what his floor would use, not on the feature-comparison sheet.
Mistake 2: Nobody on the floor was asked
The first rollout was decided in the office and handed down. Operators saw it as extra reporting with nothing in it for them, so they resisted, quietly and effectively. The second time, two supervisors helped pick and shape it. Suddenly it was "their" system.
Mistake 3: Big bang instead of phased
Going live across every department on one Monday is a recipe for chaos. The rollout that worked did one module at a time: production first, then inventory, then billing, letting each settle before the next. Slower on the plan, much faster in practice. The same one-cell-at-a-time approach works for moving off paper job cards.
Mistake 4: No clean data to start with
Feed an ERP rubbish and people walk away from it. The first time, they imported years of messy item codes, the reports came out as nonsense, and trust died on the spot. The second time, they cleaned up their item master and BOMs first. Dull work, and the single biggest reason the system stuck.
Mistake 5: Treating go-live as the finish line
The first vendor vanished after installation. The second relationship treated go-live as the starting line. Someone owned the system in-house, questions got answered in days rather than weeks, and small tweaks kept it matching how the floor actually worked.
"An ERP doesn't fail at installation. It fails at adoption, and adoption gets decided long before go-live day."An auto-parts owner, Faridabad
Two questions decide most of this before you ever open the software: what manufacturing software actually costs, and whether you need an implementation partner at all. Get those right and most of the mistakes above never get the chance to happen.
