Why Business Can't Define Digital Requirements — And How to Fix It

The Core Problem: Business Cannot Articulate Correct Digital Requirements

A 1-billion-revenue company in East China invested heavily in digitalization: budget approved, IT team assembled, external consultants on site. The mandate: business departments must clarify requirements for needs-based development, with results in three months. At the review, IT complained that business only requested fragmented form changes, leaving core processes untouched — the output was merely an electronic ledger. Business countered that the new system was more cumbersome than manual work, and IT didn't understand their workflow. The boss was baffled: money spent, people mobilized, yet the system failed.

Three Root Causes

1. The Experience Trap — People Cannot Describe What They Have Never Seen

Humans cannot accurately describe something they have never encountered. Ask a lifelong carriage driver what vehicle he wants, and he will say "a faster horse" — never "a car." Similarly, a workshop director with 15 years of manual scheduling will list Excel formulas and cell tweaks — incremental patches to the status quo. He has never seen APS one-click scheduling boost capacity, AI inventory predictions achieve high accuracy, or IoT predict equipment failures weeks ahead. His experience anchors him to the past; he cannot invent future-state requirements.

2. Local Optimization — Every Department Optimizes for Itself

Digitalization's essence is tearing down department walls to optimize the whole value chain. But when each department submits requirements, they optimize locally. Sales wants minimal data entry — two fields, one-click order — ignoring downstream finance reconciliation and inventory accuracy. Finance demands eight-dimensional fields (customer tier, payment terms, tax rate, credit period) — ignoring sales' workload. These conflicting requirements produce a Frankenstein system: each department tolerates it, but cross-functional flow is riddled with bottlenecks. Expecting a department to request "more work for me, better for the company" is anti-human.

3. Logic Gap — Two Languages That Don't Translate

Business speaks the language of management art: "case by case," "roughly," "flexible handling." Systems speak binary logic: "yes/no," "0/1," "rules fixed, no exceptions." When they discuss requirements, they talk past each other. Business says "this client is special, price can bend a little"; the system asks "how much? what defines special? who approves? how many approval levels?" Business says "close enough, that's how we've always done it"; the system rejects a single-character mismatch. Tacit processes built on relationships, experience, and unspoken norms cannot be translated into standard data logic or code. The result is an expensive electronic notebook — paper forms moved to screens, zero value unlocked.

A Pragmatic Three-Step Path for SMEs: Small Steps, Fast Iterations

Digitalization is not a 100-meter sprint; it's an endless marathon. The practical route is eight characters: "small steps, fast iterations."

Step 1: Left Foot — Tackle One Minimum Closed-Loop Scenario

Don't run menu-style surveys asking "what features do you want?" — that yields patches. Use diagnostic entry: find the company's shortest, most painful point. Example: inventory inaccuracy causing daily losses, overstock, and stockouts — a universal headache. Build the simplest inventory alert tool to solve just that. Many SMEs lack product managers; the factory planner, operations director, or even the boss must act as translator, turning business vernacular into system logic diagrams. Do one thing, do it deeply, close the loop.

Step 2: Right Foot — Use Tangible Results to Upgrade Cognition

Preaching a hundred times beats one live demo. Previously, a warehouse clerk checked stock by climbing three floors, flipping five ledgers, calling purchasing ten times — half a day for one answer. Now, a phone tap shows real-time stock, location, expiry, and one-tap replenishment. That shock no training or rally can deliver. Once they taste the benefit, cognition rises organically. They then proactively ask: "Can we pull in supplier lead times? Can we link production material issuance?" These new requests — born from experienced value — are the true digital requirements.

Step 3: Spiral Upward — Alternate Tool Delivery and Cognitive Leaps

Build the tool → users see real impact → cognition catches up → users propose improvements → those drive the next tool upgrade. Left foot, right foot, alternating forward. Small iterations avoid overreach and stagnation alike. Digitalization has no finish line; success is measured by how long you can keep running.

Final Thought

Business failing to state correct requirements is normal. Real requirements don't live in users' mouths or survey options. They hide in chaotic process ruins, misaligned data gaps, and the daily complaints of frontline staff. Don't rush to demand requirements. Look at the ground, take one step, then the next — the path emerges as you walk.