How a Shinagawa electrical wholesaler automated its returns process

Maruyama Denki, a third-generation electrical wholesaler operating from a compact warehouse in Shinagawa Ward, had spent decades building a reputation for next-day delivery across the metropolitan rail network. Yet as the business grew past 18,000 SKUs, one process quietly drained margin and morale: the manual handling of returned goods. Stock would arrive at the receiving bay in mixed crates, sometimes weeks after the original sale, with paperwork that often failed to match the original invoice.

The wholesaler is one of forty small and medium-sized enterprises in the ward currently working with the Shinagawa City automation program, a public-private initiative that connects local businesses with vetted technology providers and offers subsidy support to help fund implementation. Practical seminars and workshops covering factory digital transformation, IT investment, and online sales efficiency are run throughout the year, and participating firms are guided through a structured discovery, design, and rollout cycle.

What follows is the story of how Maruyama Denki tackled a stubborn back-office problem, the tools it deployed, and what electrical wholesalers in Australia can learn from the approach.

The returns bottleneck at Maruyama Denki

Returns at electrical wholesalers are rarely simple. A switchboard ordered for a Tokyo contractor may arrive with the wrong amperage rating; a bulk order of LED drivers may include a faulty batch from the manufacturer; or a customer may simply over-order to secure volume pricing and return the surplus weeks later. Each return triggered a six-step manual process: receive the goods, count them, reconcile against the original dispatch record, decide whether to refurbish, return-to-vendor, or quarantine, then finally re-shelve or write off.

The process consumed an average of 38 labour-hours per week and tied up two full-time staff members who could otherwise have been picking, packing, or supporting trade counters. Errors were common. A 2024 internal audit found that 14 per cent of returns were misclassified, which meant the business was quietly absorbing the cost of items that should have been sent back to the original supplier for credit.

A comparable challenge is well documented in Australia, where the Australian Consumer Law entitles customers to refunds, repairs, or replacements for faulty goods. Wholesale electrical distributors in Sydney and Brisbane routinely report that returns processing is the single most paper-intensive task in their operations, particularly when handling mixed orders from solar installers, commercial fit-out contractors, and the maintenance teams serving large industrial estates.

Mapping the end-to-end workflow

Before any software was selected, the wholesaler spent three weeks mapping every touchpoint in the returns journey, from the moment a customer raised a return request by phone or email through to the final disposition of the goods. The mapping exercise produced a single-page flowchart that exposed several hidden delays, including a 48-hour planning phase where returns sat in a holding area awaiting inspection.

The team also benchmarked against world-class distribution practices, drawing on guidance available through sector events such as the upcoming automation seminars listed on the Shinagawa program calendar.

Crucially, the mapping stage revealed that the bottleneck was not the physical handling of goods; it was the data layer underneath. Without a clean digital record of what had been sold, when, and under what terms, every subsequent decision was being made on incomplete information. This insight redirected the project toward barcode-driven tracking and a centralised returns database rather than a stand-alone robotic arm.

Selecting automation tools

Armed with the workflow map, the business evaluated eleven software platforms and two hardware vendors before settling on a modular solution that combined an AI-assisted classification engine with a small fleet of autonomous mobile robots for internal stock movement. The selection criteria included compatibility with the existing inventory management system, ease of integration with the local accounting platform, and the willingness of the vendor to co-fund a pilot deployment.

Several candidates were rejected because their per-transaction pricing model would have penalised a business with seasonal peaks. A particularly informative reference point for evaluating options came from a broader digital transformation perspective that framed automation decisions around long-term customer experience rather than short-term cost savings.

The chosen tools were configured to recognise the wholesaler's specific return categories — refurbish, return-to-vendor, quarantine, write-off — and to flag any item whose return reason did not match the original sales channel.

Implementation milestones

The rollout was staged over four months to avoid disrupting peak-season operations. Phase one introduced digital return-request forms that customers could complete via a simple web portal, replacing a phone-based process that often required three rounds of clarification. Phase two deployed the AI classification engine and integrated it with the barcode system used at the receiving bay. Phase three introduced the autonomous mobile robots to ferry stock between the inspection zone, the refurbishment bench, and the main shelving. Phase four focused on analytics, giving the operations manager a live dashboard of return reasons by product line, supplier, and customer.

Key elements of the rollout included:

• A two-week parallel-running period where the old manual process and the new digital workflow operated side by side.
• Weekly stand-up meetings between the warehouse team, the IT partner, and the Shinagawa program advisor.
• A short intensive training module for the two staff members previously dedicated to returns handling, who were redeployed into quality control and customer liaison roles.
• A clear escalation path for edge cases that the AI could not classify automatically, so that no return was ever silently mis-routed.

The subsidy programme covered roughly 40 per cent of the implementation cost, with the balance drawn from the wholesaler's own modernisation reserve.

Results twelve months on

A year after the system went live, the numbers told a clear story. The labour-hours required to process returns had fallen from 38 per week to 11, freeing the equivalent of one and a half full-time roles for more productive work. The misclassification rate dropped from 14 per cent to under 3 per cent, which translated into a measurable lift in supplier credit recoveries. The time taken to inspect and shelve returned stock fell from an average of nine days to less than two.

Specific gains included:

• A 27 per cent reduction in the value of stock written off as unsalvageable, because the refurbished pipeline now worked.
• A measurable improvement in customer satisfaction scores, with trade-counter customers reporting faster credit notes.
• A 12 per cent uplift in warehouse throughput overall, as the smaller trailers used to carry returns could be redeployed to outbound picking.

The most important outcome, however, was cultural. The two staff members previously locked into returns handling had been promoted into roles that used their product knowledge more fully, and morale improved across the warehouse.

Lessons for Australian wholesalers

The Shinagawa case carries several practical lessons for Australian electrical wholesalers, particularly those operating in tight inner-city warehouses around Sydney, Melbourne, and Brisbane. The Australian market is mature, fragmented, and heavily regulated by the Australian Consumer Law, which means that returns documentation is not optional — it is a legal requirement that protects both the wholesaler and the trade customer. Automation that produces a clean, auditable trail of every return is therefore not a luxury; it is risk management.

Australian wholesalers should resist the temptation to automate physical handling before automating the data layer underneath. The Shinagawa example showed that a barcode-driven workflow and a centralised returns database delivered most of the benefit, with mobile robots adding incremental gains on top. Vendors in Australia, including those aligned with NECA and Master Electricians Australia, increasingly offer modular solutions sized for single-site operators rather than national chains.

There is also a discipline lesson here that applies well beyond returns processing. A structured, repeatable approach — much like a flat betting approach in any high-variance activity — outperforms a clever-but-inconsistent one. Set the process, follow the design, measure the outcome, and adjust the inputs.

If your business is considering a similar digital transformation, the Shinagawa program team can connect you with vetted local technology partners, walk you through the subsidy application, and invite you to the next round of practical workshops. Reach out through the events page to reserve a place at an upcoming seminar or to request a one-to-one consultation with an automation advisor.