Every other week the same WhatsApp lands. 'Bro, I nak buat platform macam Udemy, but for my own courses. Berapa ya?' Or the e-commerce version, usually from someone who just looked at their marketplace statement: 'We want our own website, tak nak bagi commission to Shopee already. Can or not?' Both are fair questions. Both get the same lazy answer from most people, which is 'depends'. So this is the long version of 'depends': what an LMS and an e-commerce system actually cost in Malaysia in 2026, how long they really take, and the five things that always kena somewhere around month two.
We build websites and apps out of Setia Alam, and a decent chunk of that work is exactly these two systems: course platforms for training companies and coaches, and online stores for brands who are tired of paying a marketplace to own their own customers. The numbers below are the ones we quote and the ones we watch other Klang Valley studios quote. No inflate, no sandbag. If your budget is smaller than the band, we would rather tell you here than on the third call.
LMS and e-commerce look same-same, but different
On paper they are cousins. Both need accounts, both need to take money, both need content that somebody has to keep updating. That is why clients often assume one is basically the other with different words on the buttons. The difference is in where the project gets hard, and the two get hard in completely different places.
| System | The easy part everyone demos | The part that actually kills projects |
|---|---|---|
| E-commerce | Product page, cart, checkout | Stock accuracy, courier rates, refunds, and keeping it honest with your accounts |
| LMS / course platform | Upload video, take payment, done | Video bandwidth bills, piracy, progress tracking, and students who ghost at lesson three |
| Both in one (membership plus shop) | Sounds efficient on paper | Two different journeys sharing one login, which roughly doubles your edge cases |
That last row deserves a warning. Clients love the idea of one system that sells physical products and courses and memberships, all under one login. It is doable, we have built it, but understand that you are not buying one system with a bonus. You are buying two systems that have to agree with each other about who a customer is, and that agreement is where the bugs live.
Three ways to build, and only one of them is 'custom'
Before anyone talks price, work out which of these three you are actually shopping for, because people use the word 'custom' to mean all three and then wonder why the quotes are RM3,000 apart or RM30,000 apart.
| Route | Roughly what it costs | Good for | Where you will feel the ceiling |
|---|---|---|---|
| Ready-made SaaS (Shopify, Teachable, Kajabi, Thinkific) | RM150 to RM600 a month, plus a cut of transactions | Testing demand, your first hundred customers, solo operators | Local payment methods, BM interface, and anything your business does its own way |
| Plugin build (WooCommerce, LearnDash, TutorLMS on WordPress) | RM8,000 to RM20,000 once, plus RM2k to RM6k a year upkeep | Most Malaysian SMEs, honestly | Speed past a few thousand products or students, and plugin conflicts on update day |
| Custom build | RM20,000 to RM60,000 and up | Businesses whose workflow is the product | Nothing much, but maintenance is now yours forever |
Read that table bottom-up, not top-down. We have talked more clients out of a custom build than into one, and it is not because we don't want the work. It is because maybe six out of ten businesses asking for custom would have been served perfectly by a plugin build with a proper design sitting on top of it, and the RM25,000 they saved would have done far more good in photography, content and ads. Custom earns its keep when your workflow is the thing people are paying for. If your workflow is 'sell product, ship product', WooCommerce sudah cukup lah.
E-commerce: what the ringgit actually buys in 2026
- RM0 to RM7,000 a year: Shopify or a WooCommerce theme you set up yourself. One gateway, standard checkout, template design. Correct answer if you have not proven that people will buy from you outside a marketplace yet.
- RM8,000 to RM18,000: a proper WooCommerce or Shopify build. Real design instead of a theme, one payment gateway, one courier integration, product data structured properly, SEO foundation from day one. This is where most Malaysian SMEs should be.
- RM18,000 to RM45,000: custom storefront or headless build. Multiple gateways, marketplace sync with Shopee, Lazada and TikTok Shop, integration with SQL Account or AutoCount, customer accounts with order history, performance tuned for mobile.
- RM45,000 and up: multi-warehouse stock, B2B pricing tiers and credit terms, ERP integration, custom logistics logic, or a store that has to run across several brands or countries. Only makes sense when the operation is genuinely that complicated, not when you hope it will be one day.
One number to keep in your head while reading that list: a marketplace typically takes somewhere in the mid-single digits to low teens of a percent once you add commission, payment fees and the campaigns you have to join to stay visible. Own-site economics only beat that if you can actually bring your own traffic. If you have no email list, no content and no ad budget, a beautiful RM30,000 store will just be a quieter version of your Shopee shop. Build the store, but budget for the demand too.
LMS: what the ringgit actually buys
- RM0 to RM6,000 a year: Teachable, Thinkific, Kajabi, or a Circle community with payments. Video hosting included, no maintenance. Best first move for a coach or trainer with under a few hundred students.
- RM8,000 to RM20,000: WordPress with LearnDash, TutorLMS or LifterLMS. Local payment gateway, certificates, quizzes, drip content, basic reporting, and a design that doesn't look like a plugin demo.
- RM20,000 to RM60,000: custom LMS. Cohorts and intakes, granular roles for trainers, admins and corporate clients, proper progress and attendance reporting, bulk enrolment for company clients, and integrations with whatever your ops team already uses.
- RM60,000 and up: multi-tenant or white-label, where each corporate client gets their own branded space, plus SSO, deeper analytics, and an SLA. At this point the platform is the business, and it should be budgeted like one.
The hidden line item in every LMS is video. Nobody quotes for it and everybody pays for it. Self-hosting course videos on the same shared hosting as your website works fine for about a week, then falls over the moment thirty people press play at once. Realistic answer is Bunny Stream, Cloudflare Stream, Vimeo or similar, which for a working Malaysian training business usually lands somewhere between RM100 and RM500 a month depending on watch hours. Cheap compared to the alternative, but it is a monthly bill nobody warned you about.
The payment gateway conversation nobody has early enough
This is the single most common place we see a Malaysian project lose money quietly. Somebody builds a lovely store, wires up Stripe because the developer knew it, launches, and then wonders why conversion is terrible. The reason is simple: a lot of Malaysians pay by FPX online banking, and if your checkout only offers cards, you have politely asked a big slice of your market to go away.
- FPX online banking through a local gateway: iPay88, Billplz, ToyyibPay, senangPay, Curlec or Razer Merchant Services. For most Malaysian stores this is not optional, it is the default rail.
- E-wallets: Touch 'n Go eWallet, GrabPay, Boost, ShopeePay. Cheap to add through the same gateways and meaningful for lower-value baskets.
- Cards: Stripe or your gateway's card rail, for corporate buyers, foreign customers, and anyone paying for a bigger ticket course.
- Instalments and buy-now-pay-later: worth considering above roughly RM500 a basket, especially for courses, where it noticeably moves conversion.
- Cash on delivery: still real in parts of the Malaysian market, but it comes with return rates and reconciliation work, so decide with your eyes open.
On pricing: FPX is generally charged as a flat ringgit fee per transaction while cards are charged as a percentage. That difference flips depending on basket size. On a RM49 product a percentage is kinder, on a RM3,000 course a flat fee is much kinder. Rates change and gateways negotiate on volume, so confirm the current numbers directly with them rather than trusting a blog post, including this one.
The e-Invois and SST bit, since we are in Malaysia
Two things that are boring right up until they are urgent. First, LHDN's e-Invoice rollout has been phasing across businesses by size, so before you build, ask your tax agent which phase you sit in and when. Your system either needs to emit compliant invoices itself or push clean data into whatever handles it for you, and retrofitting that after launch is always more painful than building it in.
Second, SST. Do not let anyone hardcode a tax rate into your checkout. Rates and scope have moved before and will move again, and the difference between 'change one setting' and 'call the developer, wait two weeks, pay for it' is a decision somebody makes in week three of the build. Ask for it explicitly. Same goes for shipping rates, refund rules, and anything else your accountant might change without asking you first.
For training companies: the HRD Corp question
If any part of your plan involves HRD Corp claimable programmes, raise it on the very first call, not after the build. Claimable training brings its own expectations around programme registration, trainer records, attendance evidence and reporting, and the practical implication is that your LMS has to produce records that survive scrutiny months later, not just a pretty progress bar. Talk to HRD Corp about the registration side early, then tell your developer what evidence the system must be able to export. Adding real attendance and reporting to an LMS that was built without it is one of the more expensive retrofits in this whole article.
Timeline: no such thing as 'two weeks can ah?'
Transactional systems take longer than brochure sites because there is a whole category of work that only exists once money moves. Here is what a realistic schedule looks like, measured from signed brief to live, not from the first coffee.
| Phase | E-commerce | LMS | What happens if you rush it |
|---|---|---|---|
| Discovery and structure | 1 to 2 weeks | 1 to 2 weeks | Product or course data gets restructured mid-build, which resets the design |
| Design | 2 weeks | 2 weeks | Checkout and lesson screens get designed last and end up as afterthoughts |
| Build | 3 to 6 weeks | 4 to 8 weeks | Integrations get stubbed and 'finished later', which means never |
| Payment, courier and tax testing | 1 to 2 weeks | 1 week | You test in production with real customer money, which is exactly as fun as it sounds |
| Content loading and training | 1 to 2 weeks | 2 to 4 weeks | The site launches empty and sits there while somebody writes descriptions |
Add it up and a plugin build is realistically 6 to 10 weeks, a custom build 10 to 16. If someone quotes you a full transactional system in two weeks, they are either reskinning a template they already own, which is fine if they say so, or they are planning to do the testing phase using your customers. Our six-week website process is for marketing sites. Transactional builds are a different animal and we say so up front.
Five things that always kena
- Content. The build finishes on schedule and then the thing sits empty for six weeks because nobody wrote the product descriptions or recorded lesson four. Start content in week one, not after launch. This is the number one reason good builds launch late.
- Video bandwidth for LMS. Covered above, but it bears repeating because it is the bill people find out about in month two rather than in the quote.
- Stock truth. The moment you sell on your own site plus Shopee plus Lazada plus TikTok Shop with no sync, you will oversell something, and the apology costs more than the sync would have. Decide early which system is the source of truth for stock.
- Refunds and chargebacks. Write the policy before launch, build the flow before launch. Handling your first refund manually through WhatsApp and a bank transfer is a rite of passage nobody needs.
- Who runs it after go-live. Every system that quietly died in our experience died because no named person owned it. Not a department, a person, with time actually allocated in their week.
When you should not build your own
Genuinely, some of these projects should not happen yet, and a studio worth hiring will say so before taking your deposit rather than after.
- You have not sold the thing yet. Sell ten courses or a hundred products manually first. Manual is ugly, but it teaches you what to build, and the lesson is much cheaper than the build.
- Your entire traffic plan is 'we will do SEO later'. A store with no demand plan is just an expensive brochure that also has a cart.
- Nobody internally has time to run it. If your one operations person is already at capacity, adding a platform adds work, it does not remove it.
- The budget only covers the build. If there is nothing left for content, hosting, gateway fees and the first round of fixes, you are launching a system with no fuel in it.
What a proper build should include, no matter who builds it
- A staging environment. Changes get tested somewhere that is not your live store on a Friday afternoon.
- Backups with a tested restore. Untested backups are a feeling, not a safety net.
- Clear user roles. Admin, staff, trainer, customer, each with permissions that make sense.
- Page-level SEO control built in, not bolted on as a plugin later. Same principle we cover in our CMS guide.
- Mobile-first checkout and lesson pages, tested on real mid-range Android phones on mobile data, not just on the designer's laptop.
- Analytics and conversion tracking wired at launch, so you can tell what is working in month one instead of guessing.
- Handover with everything in your name: domain, hosting, payment gateway account, analytics. Non-negotiable, and one of the questions we tell people to ask in our guide to choosing an agency.
- A training session for the people who will actually use it, plus a warranty window for fixes after launch.
Honest summary
For most Malaysian SMEs in 2026, the right first move is not a custom platform. It is a ready-made tool to prove demand, then a well-designed plugin build in the RM8,000 to RM20,000 range once the numbers justify it, then custom only when your workflow genuinely is the product. The businesses that regret their spend are almost never the ones who started too small. They are the ones who bought RM40,000 of system to serve a demand they had not built yet, and the same trap we describe in the real cost of a cheap website works in reverse at the expensive end too.
If you want an honest read on which band you are actually in, including the version where we tell you to stay on Shopify for another year, we do a free 30-minute call on Google Meet. Bring your current sales numbers, your traffic if you have any, and who on your team would run the thing. We will tell you what we would build, what we would not, and roughly what it costs. No deck, no pressure, and if the answer is 'not yet lah', we will say that too.
