Digitizing Your Business? The First Step Isn’t Software.

It’s not choosing software. It’s making sure someone can properly assess it — and say no.

By Jonathan JM Lee

“Can you do it?” is the first question almost every client asks me — before a single quote, demo, or case study. It’s an easy question, and the honest answer is almost always yes. That’s the trap. Winning the business usually just means giving that yes, and most consultants are happy to. The question that actually decides whether the project succeeds — should you? — is the one that rarely gets asked at all. I ask it first anyway. It has cost me easy deals; I ask it because it’s the right thing to do, and it turns out to be the smart one too.

Most digitization projects don’t fail because the software can’t do the job. They fail because of that first question everyone asks — and it’s the wrong one.

For example: a line manager decides it’s worth automating a task that eats about ten hours a month — call it RM300 to RM400 of an employee’s time. It’s tedious, so he asks the reasonable thing: “Can we automate it?”

The quote comes back: RM15,000 to RM20,000. Nobody blinks. That’s how ERP gets sold, and it’s upside down.

ERP Is Just Your Processes in Code

ERP isn’t magic — it’s your business processes written in code. Which means the software can do almost anything, so “can you build it?” always gets a yes. “Can you?” is a fake question with a fake answer. The real one is “should you?” — and nobody in the room is paid to ask it.

The ERP Automation Math Nobody Does

RM15,000 to save RM300 a month is a four-year payback — in the fantasy where the machine does 100% of the work. It won’t: the source data is often dirty, so there are false positives a human still has to check. Add the maintenance every automation needs, and the honest payback is never — the thing gets retired before it breaks even. Spending RM15,000 to partly remove a RM300–RM400 task can look like a win on the surface; there’s a saying for this kind of overkill: using a cleaver to slice tofu.

And it doesn’t even fix the real problem. The ten hours isn’t the disease — the dirty data is, and sometimes the clunky interface on top of it. Automating is just one option, and rarely the cheapest; the honest move is to hunt for the alternative that shrinks the problem for less. Instead you automate the mess: slow manual errors become fast automated ones that still land back on a human. You paid RM15,000 to pave the cow path.

The RM2,000 Fix That Was There All Along

Look at why it takes ten hours — the screens are clunky, too many steps. Fix the interface for about RM2,000, and ten hours becomes two.

For RM13,000 less, the cheap answer saves more time and pays back in months. Both options existed the whole time. So why does the expensive one win?

How a Biased Proposal Reaches the Boss

Listen to how the request reaches the boss:

“That reconciliation task my team does? We’re still doing it by hand — ten hours a month, and we keep catching errors after the fact. A vendor can automate the whole thing. One-time RM15k and it’s done. Frees my people for real work, cuts the errors. It’s 2026 and we’re still doing this manually.”

Not a word of that is a lie. But the RM300 never gets said. The solution is pre-decided. The cost is dressed up as a one-time purchase. The benefit is vague and unarguable. “It’s 2026” makes saying no feel like being against progress. It even promises to cut errors — the one thing automating dirty data will fail at. The employee is completely sincere. That’s what makes it deadly.

Because every incentive in the room points at the big number:

• The department wants the automation — it doesn’t pay for it, and “we automated it” is a trophy; “we fixed a screen” is invisible.
• The consultant wants the automation — his cut scales with the deal, so he won’t volunteer the RM2,000 fix, or the maintenance and false-positive costs that come after. If he’s there to sell, that’s not a lapse; it’s the job.
• The boss hears a sincere pain-pitch, a confident expert, and no alternative — so he signs.

The RM2,000 answer — better on every axis — has no advocate. Nobody’s paid to propose it, nobody’s flattered by it. The best answer has no salesman. The market doesn’t pick the right answer; it picks the one the most people in the room profit from or feel good about. That’s almost never the cheap one.

Best Practice: Create a Trustworthy Central Authority for Governance — First

A judge panel — your central authority for governance — is a small, cross-functional group with the real authority to make and enforce the hard calls. Not a monthly meeting that rubber-stamps requests — a body that can tell a department no and make it stick, that sits above any one department, and that owns “should we,” not “can we.” It exists before a single vendor is called.

Done right, here’s what it protects:

1. It creates a check and balance. A panel doesn’t guarantee the right outcome — bias, and even collusion, can still slip through; no process is bulletproof. What it does is give every request an objective second look, weighed against “should we?” and not just “can we?”, by people who don’t profit from the answer. Because the software can do almost anything, “can we?” always ends in yes — which is how a RM300 problem quietly becomes a RM15,000 quote. A group whose job is to weigh cost against value catches the cheaper alternative everyone else has a reason to overlook, and stops most of the spend that only looks like progress.
2. It removes the conflict of interest. You cannot ask the people selling the solution whether you should buy it. Even a vendor who genuinely tries to be honest — one who volunteers a real feasibility study — is compromised by the nature of the business: the assessment still comes from the party that profits if you say yes, so it can’t be fully trusted, and it may even raise eyebrows. The department is no cleaner, since it gets the benefit without paying the cost. Only the panel sits outside both incentives — the one place an honest “should we?” can actually be answered.
3. It acts as a neutral control tower. Picture the software as an airport: even a world-class one still needs a control tower to decide what gets cleared, and in what order. The judge panel is that tower. Hand it to a biased controller and the wrong things get waved through — a quiet “switch off the audit log” that, with no one neutral to question it, simply gets done; a runway reserved for a favoured vendor — while better, cheaper proposals are left circling and never reach the boss. A neutral panel clears every request on merit, so the person signing sees the full picture, not just what a biased controller chose to let through.
4. It governs who can see and do what. Most ERP platforms offer very granular access controls, managed by the client alone or jointly with the vendor — but that granularity is only as good as the policy behind it. Whoever controls access controls what everyone else can see and change, which makes it power, not paperwork. Leave it in the wrong hands and it quietly gets shaped around one person’s interests: the owner’s view narrowed, a manager’s own privileges widened, everyone else boxed into whatever suits the person holding the settings. Without a neutral panel owning the policy, no one has the standing to enforce it — the settings simply drift toward whoever happens to be configuring them. It’s a governance decision, not a technical one to hand downward. You simply can’t leave it to a biased party. Full stop.
5. It builds participation into key approvals. The actions that matter most — releasing a payment, approving a budget, signing off a request or a high-value change — should never rest on a single person’s click. Most ERP systems can enforce multi-step approval workflows, but someone has to decide who participates and where each sign-off sits. Leave that to an interested party and the approval chain quietly bends to suit them; set it with a neutral panel, and every critical function carries the right checks — so no one hand can push it through alone.

Its uncomfortable job is to say no to the very people it serves — and that’s exactly why it works.

The Honest First Question Before You Buy ERP

So the honest first question isn’t “what’s your budget?” It’s “who here can tell this department no?” And if the answer is nobody: You’re not ready to buy software yet. Right now you’d just be buying a very expensive way to automate your dysfunction.

That costs the easy deal. But that’s what ethical selling actually is — not a values page or a compliance line, but the willingness to lose a deal by telling the client the truth. Anyone can be honest when it’s free; the test is whether you’ll ask the question that talks them out of buying.

I’ve come to think the reason so few people in this trade are trusted is that so few will ask it.

I’d rather be the one who does.

Here at Tektician Sdn Bhd, we’re glad of any chance to help your business grow — and, like everyone, we like to make a profit. But we believe a failed project hurts both parties far more than any profit we might reap from it. Your success and ours are the same thing. We’re not selling you the best software — we’re selling you the opportunity to build it right, from the foundation up.

Before your next ERP quote, ask your team one question: who in the room is empowered to say no? If you’d like a neutral second look at your digitization roadmap before you commit, let’s talk.