← All articles

Blog

A Marketplace Seller's Odoo Roadmap: The First 90 Days

A plain 90-day plan for Indonesian marketplace sellers adopting Odoo: what to do each month, what syncs, what goes wrong, and when to bring in help.

6 min read
  • mid
  • odoo
  • marketplace

Short answer: for a seller on Shopee, Tokopedia, TikTok Shop or Lazada, a realistic first 90 days with Odoo is not “install and sync everything”. It is three stages: month 1 to define and clean (products, stock, processes), month 2 to configure and test on real orders, and month 3 to go live in a controlled way and stabilise. Sellers who try to connect every channel on day one usually spend the money and still end up back in spreadsheets.

This is a planning guide, not a promise of a fixed timeline. Your catalogue size, number of channels and team habits will move the dates.

Before day 1: decide what Odoo is for

Pick the two or three problems you are paying for today. Typical ones for marketplace sellers:

  • Stock counts that differ between channels, causing oversells and cancellations.
  • Orders re-typed by hand from several seller centers.
  • No reliable view of profit per product after fees, discounts and returns.
  • Invoicing and tax work done separately from sales.

Write these down with a number next to each (hours per week, cancellations per month). You will use them in week 12 to judge whether the project worked.

Days 1 to 30: define and clean

Implementation guides tend to follow the same shape: planning, configuration, testing, go-live, then support. One public Odoo implementation guide (written for a bakery example) describes five phases, with master data imported during configuration and current inventory counts and open invoices imported at go-live. Month 1 is the discovery part, and for marketplace sellers it is mostly about data.

1. Fix your product master. The same item often has different names and SKUs on each marketplace. Decide one internal SKU per sellable item and map each channel listing to it. Bundles and variants (size, colour) need explicit rules. This is dull work and the single biggest cause of failed syncs later.

2. Count your stock. Do a physical count. Opening stock in a new system is only as good as the count it starts from.

3. Map the order journey. Write the path of one order: marketplace notification, packing, courier pickup, delivery, return or complaint, payout. Note who touches it at each step. Odoo should follow your real flow, or you must consciously change the flow.

4. Decide the connection approach. Options include a packaged connector from the Odoo Apps store, a third-party suite, middleware, or a custom build. Third-party listings exist for Tokopedia and, from vendors such as IZI, for Shopee, but coverage, supported Odoo versions and price differ by vendor. Treat vendor listings as marketing until you have seen a demo on your own order types. Coverage of other channels such as Lazada and TikTok Shop varies, so confirm each channel with the vendor.

Days 31 to 60: configure and test on real orders

Configure the core first. Inventory, sales, purchasing and accounting should work for one channel before you add more. A common sensible order is: your largest channel first, the others after it is stable.

Handle Indonesian tax and invoicing early. Odoo’s Indonesian localization documentation describes e-Faktur and a Coretax export module, and notes that company and customer tax details (such as NPWP and PKP status) must be set up for e-Faktur to be generated. If you issue tax invoices, ask your accountant to confirm the setup against current Directorate General of Taxes requirements, because tax rules change.

Test with real, messy orders. Run a week or two of parallel orders and include the awkward cases:

  • Cancelled orders and partial shipments.
  • Returns and refunds, including courier-rejected parcels.
  • Orders for bundles and out-of-stock variants.
  • Marketplace fees, shipping subsidies and voucher splits, which determine the real margin.
  • Fulfilment through a marketplace warehouse, where stock sits outside your own.

Know what syncs and what usually does not. Typical flows are orders in, stock levels out, and product updates. Things sellers often discover do not sync fully, or only partly: some buyer personal data (marketplaces limit what sellers receive), exact fee breakdowns, and platform-managed fulfilment stock. Ask each vendor, in writing, what is covered for each channel.

Train in roles. Warehouse staff, customer service, finance and the owner need different sessions. Short role-based training with real orders works better than a long general demo.

Days 61 to 90: go live carefully and stabilise

Pick a go-live date outside your peak. Avoid double-date sale events and Ramadan or year-end rushes. Marketplace campaigns bring order spikes that expose every weakness.

Go live in steps. A phased launch is a common-sense way to reduce risk. For a seller that can mean one channel first, then a second channel two or three weeks later, with the old spreadsheet kept read-only as a fallback for a short period.

Reload final balances. Do a final stock count and load open balances at cut-over, in line with the final data-migration step in the implementation guide above.

Watch daily for the first weeks. Check each morning: stock differences between Odoo and each seller center, orders that failed to import, and shipping statuses. Assign one named owner for this.

Review against your starting numbers. In week 12, compare with the figures you wrote down before day 1. If cancellations and manual hours fell, continue adding channels. If not, find out why before spending more.

What goes wrong

  • Dirty product data. Mismatched SKUs cause oversells. Fix this before you connect anything.
  • Platform changes. Marketplaces change their seller tools and access rules. For example, Tokopedia sellers were moved into a combined Tokopedia and TikTok Shop Seller Center during 2025, and the Trade Ministry said in June 2025 that, technically, the move did not breach the e-commerce trade regulation, according to Bisnis.com. Verify the current arrangement in your seller center, and ask any connector vendor how quickly they follow such changes.
  • Too much customisation. Every custom change adds cost at upgrade time. Use standard features until they clearly fail you.
  • No owner. Projects without one internal person who decides quickly tend to drift past 90 days.
  • Weak training. Training that is too generic or too late leaves staff working around the system.

What it costs

We cannot give a reliable single figure, because it depends on Odoo edition and hosting, connector licences (the two listings cited here were priced at roughly USD 480 and USD 1,430 when checked, so check the current price), consulting time, data cleaning effort and training. Ask for a quote that separates one-time costs from recurring ones, and ask what happens at your next Odoo version upgrade.

When to get help

Consider outside help if you sell on three or more channels, use marketplace-run fulfilment, have thousands of SKUs, need tax invoicing integrated, or have nobody in-house who can own the project. Help can be limited: a short scoping workshop, a data clean-up, or a review of a vendor quote can be worth more than a full outsourced build.

If you would like to talk through your channels and what a sensible first 90 days would look like for you, Teknologia Solutions is happy to have a no-charge conversation. There is no obligation, and an honest answer may be that you should start smaller.


Sources: Odoo implementation methodology guide; Odoo Indonesia localization documentation; Tokopedia Store Management listing, Odoo Apps; IZI Shopee listing, Odoo Apps; Bisnis.com, 4 June 2025. Marketplace rules, tax requirements and vendor offerings change often; verify against primary sources and your seller center before deciding.