Halter

Halter started with a bet that sounds like a joke until you watch it work: that a solar-powered collar and an app can replace the barbed wire that has run ranching for 150 years. Craig Piggott grew up on a 300-cow dairy farm in Matamata, New Zealand, spent nine months building rockets under Peter Beck at Rocket Lab, then quit in 2016 to prove cows would walk toward a gate on a sound cue sent from a laptop.

They did. Nearly a decade later the collar is on more than a million cattle across New Zealand, Australia, and the US, and in March 2026 Peter Thiel's Founders Fund led a $220M round that valued the company at $2 billion. What began as a hardware science project is now one of the largest agtech raises ever, and Halter is pushing into Ireland, the UK, and South America next.

Halter sells one system with three parts: a solar-powered GPS collar the cow wears, a tower (or now a satellite link) that keeps everything connected in the middle of nowhere, and an app where the farmer draws a fence with a finger. Cross the invisible line and the collar plays a sound cue from the correct side of the cow's head, nudging it back; ignore that and it gets a pulse a tenth the strength of a regular electric fence.

The same collars stream location, behavior, and health data 24/7, so the app also flags a cow in heat, an animal that stopped moving, or a paddock that has been grazed down. The pitch to farmers is less labor, more grass off every acre, and no posts to bang in.

Rotational grazing is the thing that makes pasture productive, and it is also the thing that grinds farmers down: stepping out polywire in the freezing dark, driving quads to shift a mob, rebuilding fences elk tore through. One Oklahoma ranch was running ultra-high-density grazing by hand and loved the results but hated what the pace was doing to them, in Kacie Scherler-Abney's words, they weren't loving what it did to the people.

Halter's answer is to make a fence a line you draw and a herd move a button you press. Ranchers report cutting a full day of physical presence down to a quick check of the app, grazing brushy ground they used to skip, and extending winter grazing because a virtual fence does not short out in snow.

Halter's customer is the person on the land: dairy farmers in New Zealand and Australia, and beef ranchers across the American West, from 400-head family operations to multi-thousand-cow enterprises. Farming covers half the planet's habitable land, and Halter frames that as its whole addressable market.

The product respects who it is selling to. Collars are cattle-only (no sheep, goats, or horses), the US support team is Colorado-based and knows ranching, and the sales pitch happens on horseback or at a kitchen table, not over a demo call. Whoever buys it has to trust it with animals they have spent generations breeding.

There is really one product, the Halter System, split by animal: a mature dairy platform refined over ten years and a newer beef line, Built for Beef, that took those lessons to US ranches. The 2026 twist is satellite: beef collars now connect straight to satellites, so ranchers can fence any terrain with no tower infrastructure at all.

Under the hood is the Cowgorithm, Halter's trademarked model of animal and pasture behavior, the layer that turns billions of raw sensor events into a single calm decision in the farmer's hand.

Halter is a hardware-plus-software subscription. The farmer owns the collars they buy, but the collars are useless without an active subscription to the Services, and Halter replaces broken hardware as part of that subscription rather than charging per repair. Contracts run for an initial term with automatic renewals and a Fee Lock Period that freezes the subscription fee until the next renewal, so pricing is per-collar, committed, and recurring. Halter does not publish prices; deals go through a local rep, which fits a product sold at the kitchen table.

Virtual fencing has several vendors (Gallagher, Nofence, Vence, eShepherd all come up when ranchers compare notes), but a University of Wyoming professor put the whole category in perspective: the tech is still at its brick-phone stage, useful today and unrecognizable in ten years. Halter's edge is that it got to scale first, with a million collars live and a decade of dairy data most rivals cannot match.

The risk ranchers name is the same one every challenger faces: the collars depend on connectivity that rural land does not always have, and the upfront cost is real for anyone who already sank money into wire.

The traction is concrete: more than 2,000 farmers and ranchers, over a million collars sold, and, by Halter's count, US customers drawing 11,000-plus miles of virtual fence worth roughly $220M in avoided conventional fencing.

The per-farm stories are the real proof. An Oklahoma ranch runs two to three times its county's average stocking rate on Halter and takes 60% or more of its forage per pass instead of 30-50%. A New Zealand farmer reports saving three hours of labor a day. Independent research from the Tasmanian Institute of Agriculture found no rise in cattle stress versus traditional methods, which matters when your buyers love their animals.

The recurring praise is about time and calm: mornings that start with a glance at the app instead of a dawn fence run, herds that move at their own pace instead of bunching and charging the wire, and access to brushy or steep ground that was never worth fencing before. Ranchers repeatedly describe the moves as more mellow and the cattle as less stressed.

The recurring gripes are cost and connectivity. The most common objection reps hear is the upfront investment, especially from people who already spent years building physical fence, and anything that leans on cell or wifi coverage can get flaky across wide-open country. Early adopters also note the usual new-tech glitches: collars occasionally falling off, software still maturing.

In March 2026 Founders Fund led a $220M Series E at a $2 billion valuation, one of the largest agtech rounds ever, with Blackbird, DCVC, Bond, Bessemer, NewView, Ubiquity, Promus, and Icehouse Ventures all back in. That doubled the roughly $1B valuation Halter carried after its $100M Series D just nine months earlier.

The cap table has been unusually blue-chip for a New Zealand farm-tech company since the start, in part because Rocket Lab's Peter Beck wrote an early check and took a board seat when his former employee left to chase cows. Notably, Halter now argues it is not really an agtech company at all, it is a robotics and AI company that happens to run on livestock.

Halter's stated ambition is not modest: it wants collars on a meaningful share of the world's cattle and frames farming's half-of-habitable-land footprint as the prize. The near-term plan funded by the Series E is geographic, new field and commercial teams across its three current markets plus launches in Ireland, the UK, and South America, and more than 220 new hires.

The open questions are the ones the technology hasn't fully answered yet: rural connectivity, the upfront cost that slows adoption, and whether the satellite pivot makes towers obsolete fast enough to open the remote terrain where Halter's growth really lives.

Halter is loud about being a high-intensity place. Its operating principles read like a manifesto (here to win, max velocity is our default, from first principles, work hard) and its careers page literally lists the people it wants, bullshit haters and ruthless prioritisers among them. This is not your normal 9-to-5, and the company says so out loud.

It is also office-first by conviction, not habit: people who came from remote companies, Halter claims, say they could never go back. The upside for the people who thrive is a real launchpad. An engineer hired straight out of university in 2017 ran Product Engineering by 2022; a mechanical grad became a VP in six years. Career mobility here is fast if you can keep the pace.

Halter discloses pay on a chunk of its US roles, in USD, and the spread is wide because it hires both field sales and technical talent. Sales and territory roles carry the widest ranges (they read as OTE with commission on top of base), while operations and other roles sit lower and tighter. Engineering pay is set out of the Auckland HQ.

The part that applies to almost everyone is equity: Halter states its package is salary, benefits, and an employee stock ownership plan, so staff share in a company whose valuation just doubled to $2 billion.

Halter's brand marks (HALTER, COWGORITHM, PASTURE PRO, HALTER BASE, BUILT FOR BEEF) are held by Halter USA Inc., and its terms route the US software services through Halter USA LLC, while the company itself is New Zealand founded and headquartered in Auckland. The customer contract is worth reading before signing: there is no cancellation for convenience during the initial term, renewals are automatic, and hardware comes with a warranty rather than insurance, so Halter recommends farmers separately insure collars against theft, severe weather, and accidental damage.

Halter's 2026 has been a press run: Bloomberg, TechCrunch, and the trade wires all covered the Founders Fund round and the same irresistible detail, that Peter Thiel is betting $220M on AI cow collars trained on billions of hours of cattle behavior. The coverage has also started poking at the framing, with AgFunder noting Halter's insistence that it is not an agtech company at all.