Case Study ยท Manufacturing
A 1980s-Era ERP, Replaced in Eight Months
A specialty chemicals manufacturer ran three decades of production on a terminal-based system, spreadsheets, and paper batch records. Here's how it became one modern system - while the business kept running.
The Starting Point
The business ran on a menu-driven ERP from another era - the kind you navigate by typing numbered choices into a terminal. Seeing a product's container stock meant knowing the right sequence of screens by heart. There was no web access, no dashboards, and no way to ask the system a question it wasn't already built to answer.
Around the edges, the real operating system was people: customer pricing maintained in spreadsheets, batch records completed by hand on printed sheets, deliveries reconciled against orders manually. The knowledge was real - decades of it - but it lived in heads, paper, and workarounds.
And some questions simply had no answer at any price: Can we fill this order by Friday, given the tanks, the containers, the production schedule, and every other order ahead of it? The old system had no concept of that question.
What Changed
The manual work the new system took over.
Before
Checking whether an order could be filled meant reading tank levels, counting containers, checking the production schedule, and knowing which other orders were ahead in the queue.
Now
Every order line now shows availability the moment it's entered - computed in about a second, with a full breakdown of where the stock is coming from and which orders are ahead of it.
Before
Customer pricing lived in spreadsheets. Order entry meant looking prices up by hand.
Now
Prices fill in automatically from each customer's price list, with the full price history one click away.
Before
Batch production ran on printed paperwork - temperatures, test readings, and safety notes completed by hand on paper sheets.
Now
Digital batch sheets are generated from the product's own instruction library, record readings as production happens, and won't let a step be skipped - quality gates physically lock the next stage until the last one passes.
Before
Every sales order was re-keyed into the accounting system by hand.
Now
Orders and purchase bills push straight into the accounting system in one click, and payment status flows back automatically.
Before
Receiving a delivery meant reconciling it against purchase orders and updating stock as separate manual steps.
Now
One screen books the whole delivery - even across multiple purchase orders - updating stock, flagging discrepancies, and producing receipts as it goes.
Before
Credit exposure and customer rebates were assembled periodically, by hand, from multiple sources.
Now
Watched continuously. Over-limit customers surface on a ranked list; rebate progress is tracked against thresholds all year.
Keeping the Knowledge, Losing the Limitations
A legacy system's one real asset is that it encodes how the business actually works. So before replacing anything, we did the forensics.
The deepest knowledge was in production: thousands of step-by-step cooking instructions, quality specs, and formulas accumulated over decades. That library didn't just move to the new system - it became active. Batch sheets are now generated from it digitally, record readings as production happens, derive their safety warnings automatically from the ingredients, and enforce the sequence: a step can't be skipped, and quality gates lock the next stage until the last one passes. What used to depend on experience and paper is now built into the system itself.
And the question the old system couldn't ask - can we fill this order by the date on it? - is now answered on every order line, in about a second: tanks, containers, scheduled production, and every competing order ahead in the queue, all weighed automatically, with the full breakdown one click away.
Migration moved the entire operating memory of the business: 393 products, 425 raw materials, 115 customers, 203 vendors, 2,793 formulas, 6,050 production instruction steps, 3,335 quality specs - and nearly six years of trading history that had only ever been reachable through terminal screens, now searchable in seconds.
Why It Worked
A replacement like this is traditionally a multi-year project with a committee and a leap-of-faith cutover at the end. This one worked differently, for three reasons.
We learned the business, not just the system. Before building anything, we studied how the whole operation actually runs - sales, production, purchasing, accounting - and how it all connects. That understanding meant the new system fit the business from day one, and along the way we advised on the processes themselves: not just "here's your workflow in software," but "here's where the workflow can be better."
Their own data, from the start. We migrated the company's live data early - their real products, customers, formulas, and years of history - so everyone evaluated the new system on the business they actually run, not on demo data. When people can look up their own customers and their own orders, skepticism turns into ownership. That's what brought every stakeholder along.
Staged, never stopped. The business kept running throughout. New pieces went live one at a time and were proven on real work - staff were entering live orders and running real production batches long before the full system was finished. No big-bang cutover, no leap of faith.
Launch Was the Start, Not the End
The system runs on a maintained roadmap: refinements to the production calculators, deeper quality-data capture, expanded reporting, and new capabilities added as the business asks for them. That's the model - a system that keeps evolving with the business, instead of slowly becoming the next legacy system someone has to escape from in thirty years.
Your Business Has Outgrown Its Systems
The workarounds got you here. They won't get you further. Start with a Systems Audit and get a clear, honest map of what to keep, what to fix, and what it takes.