Case study: how a 3-branch Sungai Petani workshop chain ended the lost-job-card problem with a custom system

This is a case study. For the category-level guide — what AI actually does for a Kedah SME in 2026, where it doesn't help, and the order to deploy it in — see AI for small business in Kedah: a no-BS overview.

A 3-branch independent automotive workshop chain in Sungai Petani, Kedah was losing 8 to 12 job cards a week out of roughly 80. We didn't sell them a SaaS. We spec'd a custom job-card system, deployed it in 7 weeks, and the lost-paperwork problem went away in the first fortnight. This is what it cost, what it did, and what your workshop would pay.

The business

A 3-branch independent automotive workshop chain in Sungai Petani, Kedah. The chain is run by a husband-and-wife team with their nephew as the floor manager. They started with one branch in 2014, opened the second in 2018, the third in 2022. They service everything from Proton Saga and Perodua Myvi — the everyday Kedah car — to commercial vans for the PKNK-anchored logistics businesses in the Kuala Muda industrial belt. They do general servicing, brakes, suspension, engine diagnostics, aircond, and tyre work. They have 14 mechanics across the 3 branches and roughly 230 cars in and out per week. The chain is well-regarded locally — the owner is a certified technician who's been turning wrenches since the late 1990s, including a decade at a KL-based service centre before coming home to open his own shop.

This is the case study for the custom software side of what we do — a tightly scoped system built around the way a Malaysian workshop chain actually works, not a US-style SaaS template that doesn't fit.

The problem they were actually trying to solve

Three leaks, all caused by the same root issue — paper job cards moving between mechanics, the service advisor, the cashier, and the parts counter:

• Job cards disappeared. 8 to 12 per week out of roughly 80 across the three branches. Mechanics put a card down to grab a tool, it landed under the car, it got splashed with used engine oil, or it was still in a mechanic's pocket when the customer arrived. The service advisor then spent 20 to 30 minutes reconstructing the job from memory — usually calling the mechanic over mid-job to ask what was actually done.
• Customers argued about what was authorised. The customer sign-off on extra work was a paper signature on a hand-written card. When a mechanic added a brake-pad replacement mid-job, the customer sometimes disputed the charge 3 days later. The owner had no record beyond the mechanic's word and a faded carbon copy.
• Warranty claims were impossible to defend. A car that came back 2 weeks later with the same aircond problem had no service history beyond the printed receipt. The owner couldn't prove what was done, what parts were used, or which mechanic touched it. He refunded 4 warranty claims in the 6 months before the system went live — claims he believed were not his team's fault but couldn't defend.

The owner said on the first call: "I'm not losing sleep over the money. I'm losing sleep over the customer who says we never fixed their aircond, and we have nothing to show them but a faded carbon copy." That sentence is the brief.

What we built — and why this approach

A small custom job-card system, built around the way the chain actually runs, with four parts:

1. A tablet-based job card the mechanic opens at vehicle intake. Photos of the car, mileage, customer complaint, mechanic notes, parts request — all captured once, at the start. The chain already had Samsung tablets at the counter from a previous failed POS trial, so we built the intake form as a web app the existing hardware could run.
2. A live status board the service advisor and floor manager see on a wall-mounted screen at each branch. Every active job, colour-coded: green (intake), amber (waiting parts), red (awaiting customer approval), grey (ready for collection). The job-card pile is now a screen.
3. A customer-facing WhatsApp update that fires at each status change — "Your Viva is in the workshop. We've identified the issue. Will update by 2pm." The customer doesn't have to call to ask. We reused the AI WhatsApp concierge pattern we built for a Langkawi hotel, but stripped down to status events only.
4. A parts-issued deduction that links each job to the parts taken from the shelf, so the parts inventory is updated automatically and the service advisor can see at a glance whether parts are available for a job before promising the customer.

Why not a SaaS like AutoLeap or Shop-Ware? Because those are built for the US independent-shop market and assume a single-shop, single-mechanic-of-record workflow. The Sungai Petani chain has 14 mechanics across 3 branches, mechanics that move between branches on busy days, and a service-advisor pool. The custom system was spec'd to match — every job has a chain-level job id, every mechanic has a per-job time log, every branch sees the same queue. That's the architect, not coder difference — we studied the floor and built the system around it.

How the operation changed

Three concrete changes inside the first 6 weeks:

• Lost job cards: 8 to 12 per week → 0. Every job lives in the system. The mechanics' pockets, the service advisor's desk, the floor — none of them hold a paper card anymore. The "did we lose the job card?" question has stopped.
• The customer disputes stopped. When a customer calls back saying "I never authorised the brake pads", the service advisor pulls the job on the screen, shows the timestamped photos of the worn pads, and the customer usually says "okay, when can I pick it up". The dispute resolution time went from 20 to 30 minutes to 2 to 3.
• The morning standup got 10 minutes shorter. Before, the floor manager walked each branch at 7am to read the paper cards on the wall and figure out the day. Now the standup is a 5-minute screen review — what's amber, what's red, which mechanic is on which job. Floor-manager time recovered: roughly 1 hour per day across the chain.

The system also gave the chain something they never had: a per-mechanic productivity view. The owner can see how many jobs each mechanic closes per day, average job time, and which mechanic is slow on which job type. They're not using it for performance review yet — but the data is there, and the second-phase plan is a small business automation layer on top that surfaces scheduling and rebooking insights.

The numbers, four months in

We don't do client testimonials. We do numbers. The chain's view, four months after the job-card system went live:

• Lost job cards: 8-12 per week → 0 across 3 branches. The floor manager's reconstruction time dropped from 2-3 hours per week to 0. The chain's estimate is they were losing roughly RM 1,800 to RM 2,500 a month in mechanic time on the reconstruction work alone.
• Customer dispute resolution time: 20-30 minutes → 2-3 minutes. Dispute count went from 5-7 per week to under 1 per week across the chain. Most weeks, zero. Refund rate dropped from 4 claims per quarter to 1 in the first 4 months.
• Mechanic time-on-paper per car: 8-10 minutes → 0. The mechanic opens the job on the tablet, marks parts as fitted, marks the job complete. No more hand-written cards. Across 14 mechanics and 230 cars per week, that's roughly 35-40 hours per week of paper-time recovered.
• Customer wait time on "where's my car" calls: 4-6 minutes per call → under 30 seconds. The WhatsApp update means customers call less. When they do, the service advisor can see the job and tell them immediately.
• Repeat customer rate: estimated +12% in the first 4 months. Hard to attribute purely to the system — but the owner noticed that customers who got WhatsApp updates started bringing their second car. He also picked up 6 new fleet contracts from logistics operators in the Sungai Petani industrial belt, who specifically mentioned the WhatsApp updates during the sales call.
• Total first-phase cost: RM 28,000 to RM 55,000, spread across 7 weeks. Includes the intake app, the status board, the WhatsApp integration, the parts deduction link, and 90 days of post-handover tuning.

Payback: 4 months on the recovered mechanic time alone, 2.5 months once the dispute-resolution time is included.

What it would cost for your business

We don't publish a price sheet because the price depends on what you're running. Three honest bands for a Kedah or Perak workshop chain:

• RM 18,000 – RM 35,000 (one-off). A 1 to 2-branch workshop with up to 8 mechanics, paper job cards, and an existing counter tablet. Includes the intake app, the status board, the WhatsApp update, and the parts-issued deduction.
• RM 35,000 – RM 70,000 (one-off) + RM 1,000 – RM 2,500 / month retainer. A 2 to 4-branch chain with cross-branch mechanic scheduling, parts inventory integration, and customer history tracking. The band the Sungai Petani chain sits in.
• RM 70,000 – RM 140,000 (one-off) + RM 2,500 – RM 5,000 / month retainer. A 4 to 8-branch chain with warranty workflow, supplier order automation, and a small fleet-customer layer (corporate fleet maintenance contracts).

These are the same bands we use on the Kedah services page and the Perak services page.

What this looks like for your business

If you run a 1 to 4-branch automotive workshop in Kedah, Penang, Perak, or KL, and you're losing job cards, fighting customer disputes, or watching your mechanics spend 8 to 10 minutes per car on paper — this kind of build is the right first move. We don't sell a SaaS. We spec a custom job-card system around the way your chain actually works, deploy it in 6 to 8 weeks, and hand it over.

The next step is a one-hour call, no slide deck. You tell us how many branches, how many mechanics, what your current job-card process looks like. We tell you whether a custom build is the right answer, or whether a tighter process on your existing paper is the cheaper fix. We won't pitch you either way. Get in touch and we'll set up the call.

If your bottleneck is on the parts side rather than the job-card side, our Sungai Petani electrical retailer case study shows the dead-stock dashboard build, and our Alor Setar auto-parts shop case study shows the custom stock + order system. For the AI chatbot layer that handles the front-desk booking side, see our Langkawi resort case study.