AGRITECH & CLEANTECH
European cleantech investment averaged €2.2 billion per quarter in 2024. Early-stage deal volume fell 39% in the same period. The market is concentrating — on the companies that actually work.
AgriTech and CleanTech sit at the intersection of hardware, software, and real-world operational constraints. A demo in a controlled environment and a deployment across 200 farms or 50 energy sites are entirely different engineering problems. We assess which one the company has actually solved.
Why it’s different
The technology works in the lab. The question is whether it works in the field.
AgriTech and CleanTech companies face a failure mode that pure software companies do not: the real world. Rural connectivity, variable weather, legacy farm equipment, energy grid constraints, and uncontrolled operating environments turn promising pilots into expensive deployments. The 2025 AgTech shakeout hit hardest in CEA/vertical farming, IoT sensors, and robotics — not because the underlying technology was wrong, but because the unit economics and operational realities weren't what the pitch assumed.
01
Hardware-software unit economics look better on paper than in the field
The most common failure mode in AgriTech and CleanTech is not the technology itself — it is the unit economics. High energy costs, long payback periods, infrastructure requirements, and slow farmer or operator adoption combine to push payback timelines past the point where the investment model remains viable. A product with a technically sound architecture can still be an uninvestable business if the cost to acquire, deploy, and maintain each unit doesn't support a path to positive gross margin at realistic scale. We assess the full stack economics — not just the software margin.
02
IoT and hardware create a maintenance and reliability surface that pure-software DD misses
IoT sensors, edge computing hardware, and physical devices deployed in agricultural or industrial settings create a technical risk profile that is entirely different from cloud-native software. Firmware updates, hardware failure rates, connectivity in rural areas without reliable internet, battery life in remote deployments, and the cost of physical maintenance visits — these are not edge cases. They are operational costs that directly affect the customer relationship and the growth model.
03
Regulatory dependency is a single-point-of-failure that is rarely quantified in the pitch
CleanTech companies — particularly in energy, solar, and grid-edge technology — often have revenue models that depend on government incentive structures, feed-in tariffs, carbon credits, or EU funding mechanisms. When those incentives shift or are delayed, companies with hardware-heavy CAPEX requirements find themselves in a capital position that equity financing alone cannot support. We assess the regulatory dependency explicitly and separately from the technology assessment.
Assessment Areas
Where we focus in AgriTech & CleanTech engagements.
AI in AgriTech & CleanTech
The data from the field is valuable. Getting it into a model that works is the hard part.
Agricultural and clean energy environments generate enormous volumes of sensor data — soil moisture, crop yield, energy output, equipment telemetry. The promise of AI in these sectors is compelling. The challenge is that the data is dirty, sparse, and highly variable across geographies and operating conditions. Building AI that generalises beyond the training environment is a materially harder problem than building AI on clean, structured enterprise data.
Opportunities we verify
Yield and efficiency models trained on long-running proprietary field data. Companies that have been collecting structured, labelled agricultural or energy performance data for multiple seasons or years have a dataset that cannot be quickly replicated. Multi-year, multi-site datasets with verified outcome labels are genuinely rare and genuinely valuable. We assess whether data collection has been consistent, labelling methodology is rigorous, and models trained on it actually generalise to new farms or sites.
Predictive maintenance as a quantifiable ROI driver. In equipment-intensive AgriTech and industrial CleanTech, predictive maintenance AI has a clear, measurable return: reduced downtime, avoided replacement costs, extended equipment life. When embedded in a hardware-plus-software product with proprietary telemetry data, it creates a switching cost that pure software cannot replicate.
AI-driven energy dispatch for grid-edge applications. The European energy transition is creating significant value opportunities in grid-edge optimisation — battery storage dispatch, demand response, local grid balancing. AI systems that can forecast and optimise energy flows in real time have clear commercial applications. We assess whether the AI capability is genuinely real-time and production-ready.
Risks we surface
Models trained on controlled-environment data that don't generalise. A crop yield model trained on 50 farms in one climate region may perform poorly when deployed across three countries with different soil types, rainfall patterns, and farming practices. This is the primary reason AI-powered AgriTech products fail to scale. We test generalisation explicitly, not just accuracy on held-out training data.
Hardware failure rates that destroy customer relationships before the software moat develops. IoT hardware deployed in agricultural settings fails at rates that are hard to predict from lab testing. A product with a 5% annual hardware failure rate sounds manageable until you are doing 2,000 deployments per year with manual replacement visits costing €300 each.
Carbon credit and subsidy dependency that isn't disclosed as a revenue concentration risk. Some CleanTech business models are structurally dependent on specific policy mechanisms — carbon credits, renewable energy certificates, government installation grants. When those mechanisms are delayed, reduced, or restructured, the revenue model changes materially. We assess it as a concentration risk.
Know what you’re backing before you commit.
X-Ray delivers a full product and tech verdict on any AgriTech or CleanTech target in one business day — covering the hardware-software integration, the unit economics at scale, the data quality, and the regulatory dependency.
250+ European engagements · 100% partner repeat rate