Every SME owner asks me the same question eventually: should we sell through Shopee and Tokopedia, build our own website, or just run the whole thing through Instagram and WhatsApp. The honest answer is that the website vs marketplace for selling online debate is not a single decision, it is a portfolio decision, and most businesses in Indonesia need at least two of the three channels running at once.
I have seen owners burn six months building a beautiful website nobody visits, and I have seen owners stay 100% marketplace-dependent until a platform policy change or a fee hike wiped out their margin overnight. Both mistakes come from treating channel choice as a taste preference instead of a numbers problem.
The framework I use with clients is simple: look at your margin per transaction and your repeat-purchase rate, then decide how much of your business you can afford to rent versus how much you need to own.
Marketplaces rent you customers
Tokopedia, Shopee, and similar platforms give you instant traffic, trust infrastructure, and payment handling. For a new product or a business with low repeat-purchase rate (think: a one-time gift item, a seasonal product), this is close to free customer acquisition. You should be there.
But you are renting, not owning. The platform:
- Takes 2-5% commission plus ad spend to stay visible in search
- Owns the customer relationship (you rarely get an email or WhatsApp number)
- Can change ranking algorithms, fee structures, or promo rules without warning
- Puts you next to five competitors on the same page, one click away
For low-margin, low-repeat products, this trade is fine. The cost of acquisition is baked into the commission, and you were never going to build a direct relationship with a customer who buys once a year anyway.
Your own channel owns the customer
A website (or a well-run WhatsApp Business + own payment flow) is where you build an asset instead of renting traffic. The upside only shows up over time:
- You capture the customer's contact and purchase history
- You can retarget, upsell, and run loyalty programs without platform permission
- Margins are higher because there is no commission, only payment gateway fees (typically under 3%)
- You are not one algorithm change away from losing visibility
The catch: you have to generate your own traffic. A website with no marketing behind it is a business card, not a sales channel. This only pays off if you have a repeat-purchase business (a coffee subscription, a skincare brand, a B2B parts supplier) or a high enough margin per sale that even slow organic or referral traffic justifies the investment.
Social commerce is the middle ground, not the endpoint
Selling directly through Instagram DMs or WhatsApp catalogs works well for discovery and for businesses run by one or two people. It is low cost to start and matches how a lot of Indonesian consumers already shop. But it does not scale operationally: manual order taking, no real inventory sync, and no structured customer data past a chat thread.
Treat social commerce as your top-of-funnel and lead capture, not your permanent backend. The businesses that get stuck here are the ones still manually copying orders from Instagram comments into a spreadsheet after they have outgrown that stage.
A channel-fit framework by margin and repeat rate
| Repeat-purchase rate | Margin per sale | Recommended channel mix |
|---|---|---|
| Low (one-time buy) | Low | Marketplace only, minimize ad spend |
| Low | High | Marketplace for reach + own site for high-ticket buyers |
| High | Low | Own channel is mandatory, marketplace as discovery only |
| High | High | Own website/app first, marketplace as secondary, social commerce for acquisition |
If you don't know your repeat-purchase rate, that is the first number to go find before you argue about channels. I've had this exact conversation shift in five minutes once an owner pulled up their own order data and realized 40% of revenue came from repeat customers they had no direct way to reach outside the marketplace inbox.
Plan the migration, don't force a switch
The businesses I've helped move successfully never abandoned marketplaces cold turkey. They ran a deliberate migration:
- Keep marketplace as the acquisition engine.
- Add a lightweight own-channel (WhatsApp catalog or simple website) purely for repeat customers, incentivized with a small discount for switching.
- Track what percentage of repeat revenue has migrated over 3-6 months.
- Only invest in a full e-commerce build once that own-channel revenue justifies the engineering cost.
This mirrors how you'd approach any process mapping before automation: understand the actual flow and volume before you build infrastructure for it.
The takeaway
Don't pick a side in the marketplace-versus-website debate, build a portfolio that matches your margin and repeat-purchase reality. If most of your revenue is one-time, low-margin sales, marketplaces are doing their job, stop feeling guilty about the commission. If you have real repeat customers and decent margin, every month you delay building an owned channel is a month you are paying rent on relationships you should already own. Start by pulling your actual repeat-purchase numbers this week, that single data point will settle the argument faster than any strategy deck. If you want a second opinion on where your business sits on that table, that is exactly the kind of conversation worth having with a tech partner.