.webp)
When You Can't Raise · Part 3 of 3
When You Can't Raise
- How AgTech Startups Grow When Venture Funding Dries Up
- What Non-Dilutive Funding Is Available to AgTech Startups?
- When Should a Startup Cut Costs Instead of Chasing Revenue? (current)
Cut costs when the runway you have is shorter than the time it will realistically take to reach your next funding event. That is arithmetic, not sentiment, and the comparison is unforgiving right now: Carta reports the median gap between Series A and Series B has reached 2.8 years, the longest on record, while most companies plan against 18 months. In agriculture the same comparison has a harder floor, because revenue arrives in a seasonal window: a company that missed this season's ordering window is not two quarters from revenue, it is two quarters from the next chance to sell, and further still from being paid. Cut once and deeply rather than repeatedly, because the evidence across industries is consistent that serial small cuts do more damage than a single decisive one. And cut while cash remains, because a company with six months left still chooses its own outcome and a company with six weeks does not.
Which ag and food tech companies are running out of money?
Enough of them, at a scale the sector has not seen before. PitchBook's early count has agtech deal activity down sharply in the first half of 2026, which Part 1 of this series covers in detail.
Adam Bergman of EcoTech Capital put the consequence plainly in March 2026, predicting that "2026 there will be the largest number of agtech and foodtech bankruptcies, restructurings, and fire sales to date". His accounting of the sector is worth sitting with: of roughly $40 billion raised into agtech and foodtech historically, he estimates $8 to $10 billion went into companies that went bankrupt, while only about $2 billion has been returned to investors.
The named cases are not marginal companies. Bowery Farming, once valued at $2.3 billion, ceased operations with 48 hours of notice and laid off 187 people. Plenty peaked at a $1.9 billion valuation in January 2022, recapitalized at under $15 million in January 2025, and filed Chapter 11 that March with existing equity wiped out. Believer Meats, which raised more than $390 million for cultivated chicken, simply ran out of cash in December 2025. FarmWise, an autonomous weeding company, closed in April 2025 citing both the funding environment and pressure on farmer purchasing decisions.
One caveat that sector coverage usually skips. Bowery, Plenty, and Believer Meats were capital-intensive businesses whose core problem was cost per unit, and no amount of sales discipline closes a gap that large. FarmWise is the more instructive case for most readers: an equipment business with real customers that closed when the funding environment and grower purchasing power moved against it at once.
The more relevant failures are the ones missing from that list entirely. Software and precision ag companies rarely file for bankruptcy. They get absorbed in acquihires that never get announced as distress. That is an informed read rather than a measured one, since those transactions are private and no one publishes the denominator, but it points to actual attrition in the categories most readers occupy running higher than the headlines suggest.
Capital raised is not protection. In several of these cases it was the thing that delayed the decision.
What is the actual trigger to cut?
Most teams wait for a feeling. The trigger is arithmetic, and it takes about an hour to run.
Take the months of runway you have at current burn. Then estimate honestly how long it will take to reach the event that funds the next stage, whether that is a priced round, a strategic acquisition, or revenue that covers operations. If the second number is larger than the first, you are already required to act, and every month you wait removes an option rather than creating one.
Then run the second test, which most versions of this exercise skip: can cutting actually close the gap? Separate what you spend on people and discretionary programs from what you spend on cost of goods, facilities, and contractual obligations you cannot exit. If burn is dominated by the second category, an opex cut buys weeks and does not change the outcome. That was the shape of the problem at Bowery, Plenty, and Believer Meats, and a company in that position is running a cost-structure decision or a sale process, not a layoff.
Be honest about the revenue side of that estimate too, because the demand environment moved against you at the same time the funding environment did. USDA forecasts farm sector working capital down 9.2% in 2026, with farm debt at a record $624.7 billion, and the Federal Reserve's July Beige Book reported agricultural conditions deteriorating on lower commodity prices, higher input costs, and tighter credit. A revenue plan built on last year's assumptions about what growers can spend is not a plan you can cut your way into.
The division takes an hour. The inputs are the hard part, because founders do not lie to themselves about arithmetic, they lie about pipeline. So constrain the inputs mechanically: count contracted revenue at full value, haircut everything else by 75%, and add six to nine months of process time on top of whatever milestone date you land on, because that is what raising takes even when it works. Use the same discipline if a sale is the event you are underwriting. A process runs six months or more from first conversation to close, and that time belongs inside the runway comparison rather than after it.
Agriculture also puts a floor under that second number that software does not have. Revenue arrives in a window. If this season's ordering window has closed, the earliest date you can book meaningful new revenue is next season, plus whatever your payment terms add after that. A company that missed the spring window is not two quarters from revenue. It is two quarters from the next chance to sell, and then some further stretch from being paid. Run your runway against that specific date rather than a generic number of months. This is routinely where a plan turns out to be short by two quarters, and it is the reason a company can be locked out of revenue for a year through no commercial failure at all.
The reason this test bites harder in 2026 is the second number moved. Carta's data puts the median time from Series A to Series B at 2.8 years, the longest interval on record, and down rounds at just over 19% of new priced rounds. A plan built on 18 months of runway to a Series B is not aggressive, it is more than a year behind the market median.
Treat 2.8 years as a floor rather than a forecast. Carta's dataset is all-sector and software-weighted, ag hardware and biology have always raised more slowly, and a median only counts the companies that did raise a Series B. The honest planning number is worse than 2.8 years, not better.
One more thing the arithmetic hides: the two numbers are not independent. A cut that touches sales or the technical roadmap extends runway and pushes the milestone out at the same time. The comparison that matters is runway after the cut against time-to-milestone after the cut has slowed you down. A reduction that buys six months and delays the milestone by four has bought two, and that is where founders who ran the test correctly still end up short.
Run the same test against the non-dilutive path in Part 2 of this series, because federal grant timelines run seven to thirteen months in a normal year and 2026 has not been a normal year.
When is pushing revenue the right call instead?
The test above returns cut for most companies in this market, which is exactly why the conditions that overturn it deserve to be named rather than assumed away. Three of them are real. Contracted pipeline: signed orders or executed contracts with delivery dates inside your existing runway, not a weighted forecast. Pre-season commitments: grower or dealer intent captured before the ordering window opens, which in agriculture is the closest thing to a forward book anyone gets. And a channel or distribution deal that pulls the revenue window forward rather than adding two quarters of integration work to it.
What those three have in common is that each one shortens the second number instead of extending the first, which is the only thing that changes the answer. If your case for pushing revenue rests on a pipeline figure rather than on one of those three, you are not choosing between cutting and selling. You are choosing between cutting now and cutting later with less cash.
Is one large cut better than several small ones?
Yes, and the evidence is unusually consistent.
Research across industries has repeatedly found that repeated rounds of layoffs damage surviving employees and organizational performance more than a single deep cut. Wayne Cascio's long-running work on downsizing and Culture Amp's employee data both point the same direction. The mechanism is not complicated. After the first cut, everyone competent updates their assumption about whether a second is coming, and the people with the most options act on that assumption first. You lose the ones you were trying to keep.
Ÿnsect is the fully documented version of the slow path. The French insect protein company raised over €600 million, then spent fifteen months in staged retreat: safeguard proceedings, judicial recovery, mass layoffs alongside a €10 million bridge, a second bridge of €8.6 million two months after that, and judicial liquidation in December 2025. Two bridges, staged reductions, liquidation anyway. Every one of those steps was defensible on the day it was taken, which is the whole problem. The pattern is a company buying time in increments too small to change anything, at rising cost, until there is nothing left to decide.
Read that as a case against increments, not as a case against bridges. A milestone-attached insider bridge is a standard instrument and frequently the correct one. What failed at Ÿnsect was serial bridges attached to no milestone that changed the company's position, against a capital-intensive biology cost base built on more than €600 million raised. One French insect protein company is not evidence about what a bridge does for a fifteen-person precision ag software business.
One caveat, since the research comes from large organizations and a fifteen-person company is not a four-thousand-person company. At startup scale a single deep cut can remove the only person who understands a system nobody documented. The principle holds anyway. It just means the order of operations matters more than the size of the number.
What has to survive the cut?
The wrong question is how much to cut. The right one is what has to be intact on the other side.
Three things usually qualify. The technical core that would be hard to rebuild, which is normally a small number of specific people rather than a department. The customer relationships generating revenue today, including the capacity to actually serve them, because a cut that leaves you unable to deliver converts a customer into a reference against you. And enough commercial capability to close the next deal, which is not the same as a full go-to-market team.
Agriculture adds a constraint software companies do not have. If growers have your product in the ground or your equipment scheduled into a spray or harvest window, the service capacity you cut is attached to somebody's crop. Support, agronomic guidance, and parts availability are not overhead during a season. They are the difference between a customer who renews and a customer telling every neighbor at the co-op what happened. In an industry this small, that conversation travels further than any marketing you were funding instead.
Everything else is negotiable, and the test is whether a line item would survive if you were starting the company today at your current size. Broad marketing programs, geographic expansion, second product lines, and anything justified by a roadmap that assumed the raise usually will not.
Size the cut to reach a defined destination with margin, not to reach the smallest defensible number. A cut that gets you to a milestone with two months to spare is a cut you will repeat.
And model the cost of cutting as part of the cut. Severance, notice periods, accrued PTO, and breakage on tools, leases, and contractor commitments typically consume one to two months of the savings before the savings start. A reduction you modeled as twelve months of new runway often nets eight, and eight instead of twelve changes the answer to the exact arithmetic in the section above.
What is the runway for?
This is the part most companies skip, and it is the difference between a restructuring and a slow shutdown.
Extending runway does not improve your position. It buys months, and months by themselves do not make a company more valuable to a buyer or an investor. Before you cut, name what will be true when the new runway ends that is not true today. A signed customer that repeats. A completed technical milestone that removes the largest risk. A cost structure that works at current revenue. Something specific enough that you would recognize it.
If you cannot name it, you are not preserving runway. You are extending the length of the failure, and the sector's own numbers show what that has cost investors.
Who else is in this decision?
Almost nobody makes this call alone. Your existing investors hold reserves they have already mentally allocated, a position on whether they would support a pay-to-play, and a view about what your cut signals to the rest of their portfolio and to their own LPs. Those three things shape the outcome as much as your burn rate does, which means the arithmetic is an input to a negotiation rather than a verdict you deliver.
Run the numbers before that conversation rather than during it, and arrive with four things: the runway math, the cost of the cut, the specific milestone you are buying, and what you need from the board to reach it, whether that is reserve dollars, a bridge, an introduction, or explicit support for running a sale process. Founders who lose control of this decision are usually the ones who show up with a problem instead of a recommendation.
Is a sale a failure?
Not necessarily. A sale that puts the technology somewhere it continues to be developed, keeps your team employed, and returns something to the people who funded you is a better outcome than most of the alternatives available at that point. What makes a sale feel like failure is rarely the sale itself. It is having no other option on the day you agree to it, and that is a function of when you opened the conversation rather than of the price.
Which is why the following matters more than it appears to. The strategic buyers most ag technology companies are counting on are managing their own downturn. Deere reported Production and Precision Agriculture net sales of $4.503 billion in its second fiscal quarter of 2026, down 14% year over year, with segment operating profit of $706 million, down 39%, and it left full-year guidance for that segment at down 5% to 10%. A corporate development team inside a segment with a 39% profit decline is not shopping aggressively. Which does not mean the door closes. It means the buyer set shifts, from strategic acquirers paying a premium to financial buyers paying a discount, which is a worse price on a slower clock rather than no price at all. AeroFarms below is what the second half of that shift looks like. Either way it is an argument for starting those conversations while you still have leverage rather than after cash forces the timeline.
AeroFarms shows what a distressed sale can preserve. It filed Chapter 11 in 2023, spent two more years in and out of distress, and was acquired by an affiliate of Palm Ventures in April 2026 in a deal that significantly reduced its debt. The platform and the facilities are still running, under an owner with a thesis. Plenty emerged from Chapter 11 in 53 days, but existing equity was wiped out entirely.
Both are outcomes. Neither was chosen from a position of strength. A company that starts the conversation with a year of runway is negotiating; a company that starts with six weeks is being processed. Acting early does not guarantee a good outcome. Acting late reliably removes the good ones.
When is winding down the right answer?
Sometimes it is, and the sector's unwillingness to write that sentence down costs founders real money.
If you have run the arithmetic honestly, cannot name a milestone that changes your position, cannot fund one with non-dilutive capital, and have no buyer conversation that leads anywhere, then the question is no longer whether the company ends. It is whether it ends on your terms or someone else's. Those produce very different results.
Agriculture also puts the decision on a calendar. You cannot responsibly wind down in the middle of a season. If growers are depending on your product for a crop already planted, or your machines are in their operating plan for a spray window, walking away in April strands real farms with real money in the ground. A founder who reaches this conclusion in October has a clean path. One who reaches it in April has an obligation first, which is exactly why the arithmetic is worth running before the season starts rather than during it.
Two questions get asked late and should be answered early. Who holds the grower's field data if the company stops operating, and who honors parts, service, and warranty obligations on equipment already sold. Those answers determine whether growers and their advisors remember you as someone who handled a hard situation well. This industry is small enough that they will remember either way.
A deliberate wind-down or an early sale does four things that running to zero forecloses. It returns something to the people who backed you. It places your team while the market still values them and while you can still make the calls. It finds a home for the technology instead of leaving it in a liquidation. And it protects the asset you keep afterward, which is your standing with the investors, operators, and customers you will want on the next thing. Founders who face this at month eighteen do better on all four than founders who face it at month twenty-four.
This is the least popular section in this series. It is also the one worth reading twice if you recognized your own numbers in the arithmetic above.
Cutting is the third path, not the first one. Before you size a reduction, confirm the other two are actually closed to you. If a repeatable sale is within reach and you have the runway to get there, Part 1 is the more useful article and cutting is not your opening move. If what stands between you and a market is a regulator, an unfinished scientific question, or a cost structure, Part 2 covers how to fund that work without giving up equity. If you have run both tests honestly and neither is open, then everything above applies, and the only variable still under your control is how early you act.
Frequently asked questions
When should a startup cut costs instead of chasing revenue?
When months of runway at current burn is less than the realistic time to the next funding event, whether that is a round, an acquisition, or revenue covering operations. That comparison is arithmetic and can be run in an hour. If the gap is negative, waiting removes options rather than creating them.
How long does it take to raise a Series B?
Carta reports the median time from Series A to Series B has reached 2.8 years, the longest interval on record, with down rounds at just over 19% of new priced rounds. Treat that as a floor rather than a forecast: the dataset is all-sector and software-weighted, ag hardware and biology raise more slowly, and a median only counts the companies that succeeded in raising.
Is one big layoff better than several small ones?
The cross-industry evidence consistently favors a single decisive cut. Repeated rounds cause more damage to surviving employees and organizational performance, because after the first cut people assume a second is coming and those with the most options leave first. That research comes from large organizations, so at startup scale apply it by deciding what must survive first, then sizing the cut, rather than picking a number and seeing what breaks.
How much runway does an ag startup need?
Enough to reach the next buying window plus the time it takes to get paid. Because agricultural purchasing clusters into a few weeks a year, a company that misses the ordering window is not a quarter away from revenue. It is a full season from the next chance to sell, and then some further period from collecting. Generic runway targets understate this, so run the arithmetic against a specific calendar date rather than a number of months.
How much should a startup cut?
Enough to reach a specific, named milestone with margin left over. Sizing a cut to the smallest defensible number produces a second cut later, which is the more damaging pattern. Decide first what must survive intact: the technical core, the ability to serve current customers, and enough commercial capability to close the next deal.
Is winding down a startup ever the right decision?
Yes, when the arithmetic shows no reachable milestone, no non-dilutive path to one, and no live buyer conversation. At that point the choice is not whether the company ends but whether it ends on the founder's terms. A deliberate wind-down or early sale returns capital, places the team while it is still marketable, finds a home for the technology, and preserves the founder's standing for the next venture. Running to zero removes all four options. In agriculture, timing also matters to customers: a company cannot responsibly stop operating mid-season when growers have a crop in the ground, which argues for running the arithmetic before the season rather than during it.
What happens to agtech companies that run out of cash?
Outcomes in 2025 and 2026 have ranged from abrupt closure, as with Bowery Farming and Believer Meats, to Chapter 11 with equity wiped out, as with Plenty, to distressed acquisition that preserves the platform, as with AeroFarms. Companies that act while cash remains retain more control over which of these they get.
Farm economy figures in this article are drawn from the USDA Economic Research Service farm income forecast and the Federal Reserve, current as of August 2026. USDA revises its forecast three times a year, so figures cited here reflect the forecast in effect at publication rather than final-year actuals.
Deciding what survives a cut is a commercial question before it is a financial one. If you are working through it, we have run this from the inside. 9 North Group embeds with ag and food tech teams as fractional operators. We don't just advise. We build. Schedule a strategy session.
For more on building the team that gets you to the next stage, see The Leadership Playbook for AgTech Growth.
