Introduction
Adoption, not functionality, is the binding constraint in local companies. What that means for how you choose systems.
The uncomfortable pattern
Ask any operations head who has been through two ERP implementations and they will describe the same arc: enthusiastic selection, a competent-looking implementation, and then a slow discovery that the shop floor is still using paper and the data in the system is six weeks stale and partly invented.
The post-mortem usually blames training. It is almost never training.
People use what they already have open
A machine operator, a delivery driver and a shop assistant have one thing in common: a phone with WhatsApp on it and no particular interest in your systems roadmap. Any process that requires them to install an app, remember a password and navigate a menu will be complied with for about three weeks.
Any process that happens in a chat thread they already check will be used indefinitely.
What this changes about evaluation
It moves adoption from a rollout concern to a selection criterion. Before comparing feature matrices, ask: what does the person who actually generates this data have to do, on what device, in what language?
If the honest answer involves a desktop browser and English, you have your answer about what the data quality will look like in month four.
The things that genuinely work in a chat thread
- Attendance check-in, geofenced and photo-verified
- Leave and regularisation requests, with approval in the same thread
- Expense claims — photograph the bill and let document AI read it
- Approvals for managers who are rarely at a desk
- Payslips and alerts, in the language the recipient reads
Measuring adoption honestly
Licences issued is not adoption. Weekly active users among the people whose work the system records is adoption. Track that number from week one and you will know whether the rollout is working long before the data quality tells you.
In summary
None of this requires a transformation programme. It requires deciding where each piece of data is created, making sure it is only created once, and letting everything downstream read from that record instead of keeping its own copy.
Vikram Chauhan, Delivery Manager at easyto.work.
