GrainSense: Choosing an Entry Mode into Southeast Asia with AUD 11 Million in Cash
GrainSense Pty Ltd is a Toowoomba-based maker of precision-agriculture soil sensors. Revenue is AUD 38 million, 62% of it from Australia, with an EBITDA margin of 16%. The board has approved a Southeast Asia entry study covering Vietnam and Indonesia, where broadacre adoption of variable-rate technology is rising.
Three routes are on the table. (1) A 51/49 joint venture offered by Jakarta distributor PT Sinar Agro: it contributes channel access and government tender relationships; GrainSense contributes IP under licence, seconded engineers and AUD 9 million of capex over three years. (2) A wholly owned sales subsidiary in Ho Chi Minh City: AUD 2.4 million, about 18 months before first revenue, full control. (3) A non-exclusive distributor arrangement: near-zero capex, but both candidate distributors already carry a German competitor's line.
Constraints are tight. Cash is AUD 11 million and the syndicated facility caps net debt at 2.0x EBITDA. The board's risk appetite narrowed after a New Zealand acquisition wrote off AUD 5.1 million in 2023. Indonesian government tenders require 35% local content, and GrainSense's sensor firmware is its only genuinely defensible asset.
1. Recommend an entry mode. Justify your choice using the integration–responsiveness grid and transaction cost reasoning, and state the two conditions under which your recommendation would change.
Read the worked answer
Frame the pressures first. Indonesia's 35% local-content rule and tender-driven demand create strong responsiveness pressure; Vietnam does not. The core sensing technology is globally standard, so product integration pressure is high. That combination points to a transnational posture, but a firm with AUD 11 million of cash cannot resource transnational presence in two countries at once. The correct move is to sequence, not to choose one posture for the whole region.
Apply transaction cost logic to the asset. GrainSense's only defensible asset is firmware IP. A 51/49 JV in which the partner controls the board and the channel, and receives the IP under licence, is the highest appropriability hazard on the table. A non-exclusive distributor is cheap but the competitor's line means no dedicated selling effort and no learning. Greenfield carries the lowest hazard but solves nothing in Indonesia's tender market.
Recommendation: a staged hybrid. Enter Vietnam first via the AUD 2.4 million greenfield sales office - cheap, fast, no local-content barrier, and it builds regional experience inside the firm. Delay Indonesia until a local-content-capable partner is validated, and when entering, take 49% while negotiating board control through reserved matters and keeping firmware on GrainSense-controlled infrastructure.
Conditions that would change the recommendation: (a) if the Indonesian tender pipeline proves to be under roughly 20% of addressable revenue, the local-content rationale collapses and a distributor suffices; (b) if contract manufacturing with a local partner is available without equity, a non-equity alliance preserves both cash and IP.
2. The board will sign the joint venture only if IP leakage and the 49% minority position are both addressed. Set out the governance and contractual architecture you would negotiate and explain the logic behind each element.
Read the worked answer
Protect the IP technically before protecting it legally. Contract clauses are enforced after the loss; architecture prevents it. Keep the firmware and the algorithm on GrainSense-hosted cloud services, with sensors authenticating to that service. The partner receives an API, never source code, and no source-code escrow. This converts an unmonitorable legal risk into a monitorable technical control.
Licence, not assignment. The licence should be limited by territory (Indonesia), by field of use (broadacre grain), non-sub-licensable, non-transferable, and terminable on change of control of the partner with an immediate revocation right.
Neutralise the 51% without owning 51%. A charter of reserved matters requiring both shareholders' consent: any IP transaction, capex above AUD 1 million, related-party dealings, changes to business scope, appointment of the finance lead, and dividend policy. Add a 3-seat board where the third seat is an independent agreed by both sides, so the partner cannot simply outvote on ordinary matters.
Stage the money. Commit the AUD 9 million in tranches gated on tender wins and gross-margin milestones, not on elapsed time. This caps downside within the AUD 11 million cash envelope and the 2.0x covenant.
Plan the exit at signing. Include a put option exercisable from year five at a formula-based fair value, tag-along rights, and a buy-sell deadlock mechanism. Avoid Russian-roulette buy-sell: against a cash-poor partner it structurally favours the deeper-pocketed party. Seconded engineers need IP-assignment and confidentiality deeds signed with GrainSense, not with the JV.
3. Set out the metrics and stage gates you would put to the board at the 12-month review, and explain why each is chosen.
Read the worked answer
The board's real question is not 'how is the JV going?' but 'should we release the next tranche?' Metrics should therefore be built around a release decision, not around activity.
Leading indicators (monthly to the JV board): tender pipeline value by stage; percentage of tenders where local content thresholds are met; number of seconded engineers still in place; partner-sourced leads converted; average selling price versus the German competitor.
Lagging indicators (quarterly): revenue in Indonesia, gross margin on sensor units, cash burn against tranche, share of revenue derived from partner relationships versus GrainSense-generated demand. That last metric matters because dependence on partner-generated demand is the warning sign that the JV is really a distribution deal with equity attached.
Integrity and control metrics: number of API access exceptions, audit log anomalies, and any partner request for source code or sub-licensing - each is an early-warning signal of appropriability behaviour and should be reported to the GrainSense board regardless of the JV board's agenda.
Stage gates: tranche two is released only if two Indonesian government-linked tenders are awarded within 12 months and gross margin holds above the plan floor. If either fails, the default is to convert to a distribution arrangement rather than to inject more capital. Stating the failure path in advance stops the board from making a second NZ-style write-off decision under sunk-cost pressure.