Digitizing a manufacturer: a 6-step roadmap that avoids multi-billion VND mistakes

The question you ask us most is where to start. The short answer is to start from last month's expense statement, not from a meeting with a software vendor. Manufacturing digitization fails at the first step more often than at the technical step, because until you know where the money is leaking, whatever you buy is the wrong purchase.

This article is the 6-step roadmap we distilled from projects for a packaging factory, a furniture workshop and a logistics warehouse chain. Each step has a timeframe, the work to be done and a metric to tell whether the step is finished. After reading, you can take it and run it, even if you don't work with MONA.

Billions of dong poured into ERP, yet the workshop still runs on chalkboards and notebooks

We see this scene in almost every factory before we start: an ERP screen hangs on the wall of the control room, with good numbers and full charts. Walk down to the workshop thirty meters away and the foreman is still writing output in a ruled notebook, team leaders still pin paper boards to the wall, and at the end of the shift someone sits down to retype it all into the system.

The system isn't broken. It is just describing a factory half a day behind reality, and that is the factory leadership is making decisions from.

Veteran staff resist quietly, and they have reasons

On the shop floor, long-serving people are an asset. They know which machine tends to jam on the third shift, which batch of raw material tends to run short, and who to call when a supplier delivers short. When the new process is finalized and only then handed down, they don't argue. They just keep working the old way and let the system run in parallel for show.

We wrote that part up as its own article on three reasons employees resist automation and how to keep your best people, because it is too long to squeeze in here. What to remember for this roadmap: if step 2 is done halfheartedly, the four steps after it can't save you.

Hiring a high-paid "digitization expert" who has never stood beside a boiler

When insiders won't change, the common reflex is to hire someone from outside, with an impressive title and a salary on par with a senior department head. That person brings a beautiful plan on slides, full of big data and forecasting models.

The problem shows in the third week, when they have to describe the process for discharging a defective batch or changing the night-shift quota. Slide knowledge doesn't reach those places. The project stalls at the survey stage, and the salary has already been paid.

Not every outsider is like this, and some do excellent work. But the selection criterion should be "has stood inside a factory like this one," not the thickness of the plan.

A multi-billion-dong ERP ends up as a place to copy Excel files into

The most common outcome is a big, expensive system where only about a fifth of the features are touched. Accounting still builds its own spreadsheets, the warehouse still keeps its own files, and at month end someone sits down to piece it all together.

MONA once took over a furniture company with 4 workshops and about 320 workers, which had spent a large sum on a foreign system and, after a year, had nearly abandoned it. We didn't suggest throwing that system out, because the money was already spent and scrapping it would be another shock for the team. Instead we built a few small modules that sit in the middle: staff keep entering data into the forms they know, and the modules read it and push it to the main system. After half a year, the share of data entering the system on time rose above 85% without anyone relearning from scratch.

The 6-step manufacturing digitization roadmap MONA is running

The six steps below go in order. Skip one and you usually have to come back and make up for it, at a higher cost.

Step 1: Trace the cash flow to find the leaks, not the process first

Most consultants open with a company-wide process survey. That takes months, costs money, and usually concludes with exactly what you already know.

We start from the money. Open the expense ledger for the last three months and find three items that have swollen abnormally: scrap, dead stock, labor hours on repeat work, and penalties for late deliveries. For a 3-workshop packaging manufacturer in Bình Dương, the biggest leak was manual stock counting and inventory reporting, enough to push the gap between the books and reality to double-digit percentages every month.

This step is done when: you can pin down exactly one step and say how much it costs you each month.

Step 2: Form a pioneer team from your core staff

Three to five people, chosen for their standing inside the company rather than for being free: the workshop foreman, the warehouse accountant, the sales manager, plus one quick young hand. This group uses it first, criticizes it first and sees the software change in response to their criticism.

At an 18-workshop packaging factory chain, we spent the first three months just sitting with and interviewing about 80 core workers before writing the first line of code. It sounds slow, but it meant the data entry screen was built close to the paper forms they had filled in for ten years, and later training time dropped to two weeks.

This step is done when: there is a list of named people, a fixed weekly meeting and someone in the group who dares to say plainly what's wrong.

Step 3: Pick one winning module to build first

A factory doesn't need an all-encompassing system right away. Pick one small battle you are sure to win: the module that hits the leak you found in step 1.

For the packaging chain above, the first module was warehouse and materials management with QR codes. It is compact, quick to build, and directly addresses the step that costs the most. A visible win in the first quarter changes the whole company's attitude toward the project, which is worth far more than building five more modules at once.

This step is done when: the module scope is written down, with clear boundaries and a line saying "this one we'll do next time."

Step 4: Run a 90-day pilot in one workshop, with metrics set in advance

The pioneer team uses the module in a single workshop or on a single line for 90 days. The metrics must be set before the start date, such as: how many percent the counting time per shift drops, how far inventory discrepancy goes down and from what level, and how many minutes it takes for reports to export automatically.

If 90 days pass and the metrics haven't moved, sit down to find the cause and adjust. Don't scale up just to keep to the schedule. Scaling something that doesn't work yet only breaks three more workshops.

This step is done when: there is a before-and-after metrics comparison table, signed by that workshop's own foreman.

Step 5: Let the pioneer team teach, with MONA standing behind them

By now the trial group has become the people who know the work best. They are the ones who train the remaining workshops, using examples from their own jobs. Words from an insider carry more weight than words from a vendor, something we have verified on every project.

MONA handles the documentation, training videos and on-call support for the first two weeks of each new workshop. If a business wants to keep this training long term instead of letting documents drift away in group chats, we build it on an internal learning platform.

This step is done when: a new workshop can come on board without anyone from MONA being there in person.

Step 6: Measure again, then continue with the next module

Go back to step 1 with the new expense ledger. Where is the biggest leak now? It has usually shifted to another step, for example from inventory to production planning or quality management.

The modules are designed to connect to each other from the start, so when the second module goes live the data flows straight through and nobody enters it twice. After three or four rounds like this, what you hold is a system tailored to your own factory, built up layer by layer rather than in one multi-billion-dong leap.

This step is done when: you have a module schedule for the next two quarters, with the expected savings for each module.

An 18-factory chain: from 4 people typing Excel to inventory reports that run by the minute

This is a manufacturing partner that operates 18 factories nationwide. Their production lines run well. The problem was in the reporting department.

Before. Four experienced staff, not low-paid, whose main daily job was opening 18 Excel files sent in from 18 factories and retyping them into one master sheet. The figures were always at least a day late, mistakes were routine, and leadership never had an accurate view of inventory at the moment a decision was needed.

What MONA did. We didn't add yet another packaged software on top of the pile they already had. Instead we built a dedicated reporting module, plugged directly into their existing production management system, pulling data straight from the production machines and the warehouse scan points. Every movement of raw materials, semi-finished goods and finished goods is recorded the moment it happens, without waiting for anyone to send an email.

After two weeks of running live. Factory directors can see real-time inventory on their phones. Errors from manual data entry dropped by about 80%. Report delay is close to zero. And the part we find most valuable: about 210 working hours a week were handed back to those four staff, who moved on to analyzing data to optimize the supply chain instead of retyping numbers.

This story illustrates steps 3 and 4 above exactly: pick one step, do it through, measure it in numbers, then scale.

The three places that break most often when a factory follows the digitization roadmap

According to McKinsey, about 70% of large-scale transformation programs fail to reach their stated goals (source: McKinsey & Company). In the factories we've worked with, most failures fall into the three places below, and all three are avoidable.

Break 1: Handing the whole project to IT while leadership stays outside

Digitization is often assumed to be a technology matter and handed entirely to IT. The IT department does very well on infrastructure, servers and connectivity. But they aren't the ones accountable for scrap targets or injection machine capacity, so the software comes out technically right and wrong for the business.

At MONA, the first session is always with leadership, not IT. We fix the business goal in numbers first, such as bringing scrap down to a set level or raising the capacity of one line, and only then design software to hit that number. For a plastic packaging chain with 4 workshop clusters, setting the goal first like this helped them cut scrap by 9% after one quarter.

Break 2: Changing the whole company in one go

An ambitious plan sounds great in the meeting room: digitize HR, accounting, production and logistics at once. On the shop floor it becomes chaos. Staff get hit by four changes at the same time, the budget balloons, and the project has no end point where anyone can declare it done.

A garment company with 3 factories in Long An did the opposite: they touched only the materials management module and nothing else. After three months inventory costs dropped 12%, and that very number is what opened the way for the next phase.

Break 3: Measuring by gut feeling, not by metrics set in advance

This is the quietest way to break. The project runs, everyone is busy, and three months later when you ask "is it better?" the answer is a vague "seems a bit better." No one can prove it, so no one dares ask for budget for the next phase.

The cure sits in step 4: write the metrics down before the day you press go, and measure again the same way after 90 days. Just three metrics is enough, as long as they can be measured and nobody can change them with words.

If you want a broader view beyond the factory, we have written about digital transformation consulting in the AI era, and about automated workflows that can replace entire departments. And if your business is torn by cheap software pitches, also read 5 mistakes of cheap automation before you sign.

Where to start this week

Three things you can do this week, with no budget needed.

Task 1: Pick the department that is bleeding the most money. Open last month's expense report. Warehouse, purchasing or the production floor: wherever there are abnormal costs, a high defect rate or inventory that has swollen for no reason. You don't need to be certain, just have a well-founded suspicion. A workshop with 80 workers that MONA once worked with started from exactly one line: auxiliary material costs up 18% in three months.

Task 2: Sit down with that department head and two people who do the work directly. No slides, no reports. Ask three questions: which repeated task takes the most time, which step most often goes wrong and has to be redone, and if a machine could take over one task, which would you pick. Write down the answers verbatim; don't polish them.

Task 3: Convert those answers into money. How many labor hours per week, multiplied by the hourly labor cost, plus the loss caused by errors. This number is what you bring to your next meeting, and also what you compare against any vendor's quote.