# Plantation OS — The ERP Question > [!important] This note supersedes an earlier position > The August 2026 pitch materials said Anthesis "does not compete head-on with estate ERPs on day one" and "sits above the ERP." The **18 Aug 2026 Utopia download call changed that.** The team now intends to build the ERP layer too. This note records why, and what it costs them. ## The problem that forced it The challenge put to the team on the call: *you may have found the right problem, but the buyer may have no base to adopt your solution onto.* A compliance layer that assumes an existing system of record only works if the buyer has one. > [!warning] There is no installed base to layer onto — and the real number is lower than first thought > The ~20% figure originally quoted is **LintraMax's own published marketing**, not an observation from the Sabah network: 700 estates against 3,939 plantation entities (17.8%), 1.0M ha against 5.70M ha (17.5%), 80 mills against 461 (17.4%). > > Three problems with taking it at face value. It is **self-reported and undated**. Its numerator spans **Malaysia, Indonesia and PNG** against a Malaysia-only denominator, so the Malaysian figure is lower. And as an enterprise system it **skews to the 2,374 entities above 500 ha** — which puts penetration inside the 40.46–500 ha target band at **likely low single digits**. > > ==The "no base to build on" argument is therefore stronger, not weaker.== For the overwhelming majority of the target band, the system underneath is pen, paper and the manager's head. ## Lintramax, as described on the call | | | | --- | --- | | **What it is** | A plantation ERP, sold as a cloud platform. Published modules cover estate operations (check-roll, transport, nursery), palm oil mill (weighbridge to lab), an online/offline field app, mapping and harvest tracking, analytics, a management app and automated attendance. | | **Onboarding** | Deep-dive discovery, DD, a proposal priced by number of manager seats, then on-site setup | | **Commercial shape** | An **annual subscription** plus a **one-off implementation charge**. The annual line is substantial — materially larger than an estate's recurring compliance spend. → [[Rimau/The Stack Not the Certificate]] | | **Published footprint** | LintraMax's own marketing: 1M+ ha, 700+ estates, 80+ mills, 200+ companies across Malaysia, Indonesia and PNG; 23 years in market | | **Hard limit** | ==Only works where there is cell signal.== Most plantations are in the middle of nowhere. | | **Real adoption** | Partial. Estates that "use" it still run pen-and-paper in parallel — a **fragmented workflow, not a unified one**. Attendance modules sit empty. | The signal-coverage constraint is the most under-rated fact in the corpus. It is not a UX complaint, it is a physical exclusion of much of the addressable market — and it argues for offline-tolerant, asynchronous capture (photos, voice notes, store-and-forward) over a live web app. ## The team's answer 1. **The ERP is not the hard build.** With AI it is a matter of weeks; Marc has already built his own systems. **The compliance layer is the hard part.** 2. **MSPO data *is* the ERP schema.** The evidence MSPO demands — how the plantation runs, how manpower is allocated, fertiliser application, land records — is precisely the data an ERP needs. Collecting it for compliance builds the operating system as a byproduct. 3. So the sequence inverts: compliance is not a layer *on* an ERP; ==compliance is how you acquire the data that becomes the ERP.== This is [[Rimau/The Flywheel and Rimau Sentient|the flywheel]] argument, applied one turn earlier than the flywheel diagram does. ## What it costs them > [!warning] The honest trade-off > - **Scope.** "The ERP is easy with AI" is the most common thing a technical founder says before it isn't. Payroll, weighbridge integration and worker tracking are unglamorous, regulated, and full of edge cases. > - **Positioning.** The deck's clean competitive story — *we don't compete with the ERP, we do the job nobody covers* — weakens the moment Rimau is also an ERP. Against Lintramax's 700+ clients they now compete on the incumbent's ground, not just above it. > - **The three conditions.** Easy, cheap, tangible. An ERP is none of those by default. Whatever gets built has to stay inside [[Rimau/Adoption Design - WhatsApp First|the adoption constraints]] or it fails the company's own test. > > **The resolution to watch for:** ship the *minimum* record-keeping that compliance requires, and let ERP surface area accrete from real usage — rather than setting out to rebuild the incumbent feature-for-feature. > [!important] The scope expansion is not reluctant — it is where the revenue is > Later evidence reframed this entirely. The ERP subscription is the **larger of the two budgets an estate already carries**, by a wide margin. So building the record layer is not a defensive necessity forced by low ERP penetration; it is the move onto the bigger line item, with compliance as the entry point. → [[Rimau/The Stack Not the Certificate]] ## The upside if it works The framing that came out of the call: the target is **the right tools that work in a way that's adoptable for owners of medium-sized plantations** — CRM, intelligence, compliance, whatever they need. Obsess over the customer, not the category. Get that right and dimensions of the market open up beyond the plantation owner. Practical shortcut raised: **don't build the CRM from scratch.** Open-source, off-the-shelf options do what HubSpot does, free. Spend the build budget on compliance intelligence and the adoption surface. ## Related - [[Rimau MOC]] - [[Rimau/Rimau Anthesis - The Wedge Product]] - [[Rimau/Adoption Design - WhatsApp First]] - [[Rimau/Competition and the Impartiality Line]] - [[Rimau/The Flywheel and Rimau Sentient]]