Sales and Distribution Management System

The problem

Growth was being held back.

Despite consistently hitting their sales targets, the company's growth was constrained by over 10% monthly customer churn, delivery delays affecting around 4% of daily sales, and over 5K liters of excess inventory accumulating.

Key insight #1

How does a company handling over ₱25M in monthly orders still lose millions each year?

After months of employee interviews & analyzing business processes, we found that customer churn, delivery delays, and excess inventory weren't separate problems — they were symptoms of the same issue.

Key insight #2

Different symptoms, one root cause: data utilization.

The data existed. It just never became information. Information never became decisions.

Management ran on gut feel because critical warning signs remained invisible until it was too late.

Solution strategy

Building wasn't the first option. It was the last one standing.

Before building anything, the real question was whether it was necessary at all. Not every business problem needs software; sometimes a policy or process change delivers greater value at a lower cost.

We evaluated multiple options against the client's multi-million-peso annual losses. Staying with the status quo recovered nothing. Manual process improvements recovered an estimated ₱1.8M annually. An integrated system recovered ₱3.8M. The size of that gap is what justified building anything at all.

Visibility, then action.

The solution is built around a two-pronged approach: First, an order management system that fully digitizes ROQSON's operations and completes the flow of data across the business. Second, business intelligence capabilities that transform this data into insights, alerts, and recommended actions.

The initial solution

We chose speed. It cost us 4 months.

We assessed multiple implementation approaches before deciding on ERPNext — an open-source platform with 88.5% requirements coverage, pre-built modules, and fast deployment. It checked every box on paper.

After multiple live testing sessions with drivers and sales staff, the same friction kept resurfacing. Processing a damaged item return or even promo eligibility checks required workarounds the system wasn't built for.

The hard call

We refused to ignore what wasn't working.

Admitting we'd chosen wrong months in wasn't easy — but it was the right call. We refused to force the client's workflow onto software it wasn't built for.

On a tight deadline, we pulled ERPNext's data models and business logic out and built a custom system around them. With agentic development doing half the heavy lifting with migration, we pivoted quickly and doubled down on rapid iteration.

Every Saturday became another opportunity to validate assumptions, uncover edge cases, and refine the system against the client's actual operations.

Fit meant sweating the small stuff.

Software should make employees' jobs easier, not harder. Every unnecessary step was removed, and every workflow was challenged. What information should already be available? What could the system do so users didn't have to?

For example, because deliveries followed planned routes, dispatchers could now simply drag and reorder trips without recreating them.

Mock data

The outcome

10 modules. 8 reporting tools. One system.

Sales knows who's about to churn. Accounting knows who's overdue. Warehouse knows what's overstocked. Dispatch knows where every delivery is in real time.

It's not a dashboard that waits to be checked. It's a system that pushes back, showing which customers to call and which stock to clear before any of it becomes a crisis. Order management, customer risk flags, demand forecasting, and delivery tracking were all built, deployed, and tested with the client's own data before handoff.

How we measured success

Each KPI mapped directly to one of the client's operational challenges, ensuring success was measured by business outcomes rather than feature delivery.

Key learnings

What we learned the hard way.