Indonesia’s agritech sector looks different in 2026 than it did when TaniHub raised its Series B in 2021. The cohort of platforms that emerged in the late 2010s — TaniHub, Crowde, iGrow, Sayurbox, and others — aimed to do for smallholder agriculture what fintech had done for SME credit: build technology infrastructure for a segment the formal banking system could not serve at scale.

Some platforms have grown. Several have restructured significantly. A few have wound down. The credit arm of the sector, in particular, has been through a harder cycle than founders or investors anticipated.

After fifteen years inside Indonesian risk functions, I have watched this cycle with direct visibility into the structural risk questions it raised. The story underneath the platform-level outcomes is more useful than the company-by-company narrative.

What the first agritech generation built

The Indonesian agritech sector that emerged after 2016 built two adjacent infrastructures. On the commerce side, platforms built marketplaces connecting smallholder farmers directly to institutional and retail buyers — restaurants, hotels, supermarkets, modern trade. On the credit side, P2P lending platforms targeted agricultural working capital specifically — seeds, fertiliser, equipment, harvest financing.

TaniHub, founded by Pamitra Wineka, Ivan Arie Sustiawan, William Setiawan, and Michael Jovan in 2016, was one of the most prominent platforms to combine both sides through its TaniFund credit arm. Its Series B in 2021 was one of the largest agritech rounds Indonesia had seen.

Also Read: Why Southeast Asian agritech must build for acquisitions, not IPOs

What turned out to be harder than expected

Three things proved structurally more difficult than early models priced in.

Harvest cycle credit timing. Agricultural loans do not behave like consumer or SME credit. They are paid back from a single harvest event that can be three, six, or twelve months out, and that may or may not arrive at the expected volume or price. A platform’s cash flow timing assumes regular monthly repayments. A farmer’s cash flow timing does not.
Default correlation. Agri-credit defaults are not independent in the way SME or consumer defaults mostly are. A drought, a flood, a pest event, a commodity price collapse — any of these correlate defaults across hundreds of borrowers in the same region simultaneously. Models that treat each loan as independently distributed misprice this correlation, sometimes by an order of magnitude.
Operational cost per loan. The cost of underwriting, monitoring, and collecting a small agricultural loan — often spread across remote geographies, often involving in-person verification — is high relative to the loan ticket. Platforms that priced loans against urban operational assumptions ended up subsidising rural credit from other revenue lines.

Lessons learned

Six principles emerge from the Indonesian agritech credit cycle.

Agri-credit is not consumer credit with mud. The underlying risk physics — harvest cycles, weather correlation, commodity volatility — is structurally different. Models built for one will misprice the other.
Correlation is the silent killer. The single biggest pricing error in early agritech credit was underestimating how correlated defaults can become inside a single weather or price event. Diversification across crops, geographies, and harvest cycles is solvency infrastructure, not a marketing point.
Operational cost is destiny. Platforms that did not build for the cost of remote, small-ticket lending from day one ended up cross-subsidising it forever — or stopped lending. There is no version of agritech credit at scale without operational cost discipline designed for the segment.
Funding tenor must match harvest tenor. Short-term retail or institutional funding does not pair well with agricultural cash flows. The platforms that survived better had funding partners willing to hold positions across full agricultural cycles, not calendar quarters.
The buyer side is more durable than the credit side. The commerce infrastructure built by Indonesian agritech — connecting farmers to buyers, aggregating produce, building cold chain — has aged better than the credit infrastructure built alongside it. Future capital allocation should reflect that.
The credit gap is durable, the model needs to mature. Indonesian smallholder agriculture will still need credit access whatever the platform-level outcomes of the first generation. The next models will need to absorb the lessons the first generation paid for, or they will pay for them again.

Also Read: Agritech investors are learning that infrastructure matters

The macro stakes

TaniHub and the agritech cohort built something Indonesia had not had before: a market-facing technology layer on top of smallholder agriculture. Not all of it survived. The parts that did, and the lessons from the parts that did not, are now the foundation for whatever comes next.

Indonesian smallholder agriculture employs tens of millions of people and produces a significant share of the country’s food supply. The credit gap at the farm gate is one of the most important development finance questions in Southeast Asia. The first generation showed that the gap could be addressed by technology platforms — and that it is harder than the original models predicted. The next generation has the data to do better. The opportunity is still there.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post The agritech credit paradox: Lessons from TaniHub and Indonesia’s first agritech generation appeared first on e27.