Archive for on-farm dairy processing

After 75 Years and 850 Doorsteps, One Number Forced Cooil’s Dairy to Choose  – How Close Are You?

Thinking about adding or expanding on‑farm processing? Read this 75-year doorstep story first. It might change your plan.

On January 31, 2026, Juan Hargraves finished the last doorstep milk delivery Cooil’s Dairy would ever make — ending an on-farm dairy processing and direct delivery operation his wife Kirsty’s family had run for more than 75 years in the south of the Isle of Man. For some customers, those rounds had been part of life for more than 60 years. Three generations of the same households opening the same door to find the same family’s milk before dawn. 

Nobody was angry. Nobody was bankrupt. The herd of 120 to 130 cows is still milked every morning. But the processing plant needed significant investment that the operation couldn’t justify, and Juan and Kirsty made the call while they still had choices—to refocus on farming and family life. “After much discussion and careful consideration,” they wrote, “we are not in a position to make this investment in the current climate”. If you’re running on-farm processing for a retail channel that only handles a minority share of your total output, the number that killed Cooil’s retail operation — their retail‑to‑wholesale ratio — is one you should know cold. 

Juan and Kirsty Hargraves with two of their six children at home on the Isle of Man. For 75 years, the Cooil’s Dairy milk round started just after 1 a.m. — but the hardest work always happened under this roof. 

Three Generations, One Route

Leslie Cooil started farming in the Port Erin area around 1942 or 1943, and the doorstep dairy that would define the family business for the next 75 years followed in the early 1950s. “The very start of it would have been Leslie Cooil in about 1942/1943, from what I can get from Ian and Gary Cooil,” Juan told Manx Nostalgia. Ian and Gary — Leslie’s sons — carried on their father’s legacy in the early 1970s, when Ian was about 21 and Gary about 5 years younger. The operation became known locally as Cooil Brothers. 

By 2010, Juan had gone from the kid who jumped on the back of the milk truck to a business partner. He first hopped on a Cooil’s truck in 1987, when he was 7, and later went up to the farm to help and worked there until he was 17. In 2004, he took in 120 acres of bare land neighboring the Cooils, running sheep and a few suckler cows while working full‑time on another farm. Buildings went up on that greenfield site in 2007, with more added over time to move the cows to newer facilities and expand the operation. In 2010, entered in to a partnership with the Cooils. In 2014, he bought Ian out upon Ian’s retirement. And in 2020, just six hours after their youngest child was born, Kirsty was signing the papers to buy out Gary’s share of the business — swapping her life as an estate agent for being fully in the dairy with Juan. All of that sits behind the one‑line summary: “Juan and Kirsty took over fully.”

“We’re Cooil’s Dairy Limited. We’ve been Cooil’s Dairy Limited since about 3 years ago now, when my wife, Kirsty, and myself took it over fully,” Juan told Manx Nostalgia in December 2023. “Obviously, our surname is Hargraves, but we’ve kept Cooil as the known trading name”.

By the time they made the decision to close, they were delivering to about 850 houses. Juan is clear: just over 1,000 would have been the peak of COVID, when they took on everyone who wanted deliveries and lived in their area. As customers went back to their usual routines — and as older clients passed away — the number settled back to roughly 850 households.

The team was small and tight. Mark ran two delivery rounds, working pretty much six days a week. Brian handled a third round on Monday, Wednesday, and Friday, and washed every bottle that came back. Lorna bottled the milk. Kirsty managed the office, the accounts, and all Farm Assurance paperwork — which Juan noted had become “a massive thing to undertake.” Juan covered the farm, the milking, and filled in on delivery routes whenever someone was off. Five people. About 850 households. Every week.

And it wasn’t just milk. Cooil’s delivered fresh Manx milk and cream in recyclable glass bottles, plus eggs, potatoes, homemade cakes, and ice cream. 

“We’ve made so many friends over the years, saved lives, moved furniture round and even caught criminals in the act,” Juan wrote in the farewell message. During the blizzard of 1994, the family later recalled, the team delivered by tractor and trailer because the milk had to get through. During COVID‑19, Cooil’s took on a wave of new customers as island residents turned to doorstep delivery — a surge that placed heavy additional demands on processing equipment already built for the existing base. 

Two Sites, One Dairy

There’s another piece you don’t see if you only watch the milk truck pull up at the door. The processing plant wasn’t even on the same site as the cows by the time the last round went out.

The herd moved to a new greenfield site in 2015, onto the newer facilities Juan and Kirsty had been building since 2007. The processing stayed at the original Cooil’s site. To bridge the gap, they retro‑fitted a DX bulk tank onto an old grain trailer chassis and hauled milk back for processing five days a week. Every load meant diesel, time, and one more moving part that could go wrong between parlour and pasteurizer.

And the work didn’t stop when the van pulled into the yard. “Every evening we had to make sure we had made any customer changes so that the rounds were ready to go just after 1 a.m., the vans were ok to go and there was enough potatoes etc. to go for the morning,” Juan wrote. Even now, a week later, with the rounds done, they’re still catching up invoices. The reality, he says, “hasn’t fully kicked in” — but they already feel a sense of freedom.

“So it fell to me to cover whatever needed doing,” he admitted. In their house, six children meant Kirsty’s hands were full, especially in the mornings. A young lad was working on the farm, but with limited experience, he couldn’t do it all on his own. Relief staff? They couldn’t afford them — and that’s if you could even find someone willing to milk cows one day, bottle milk the next, and drive rounds in the dark the day after that.

“I’ve seen it a few too many times to care to remember,” Juan wrote, “that I’ve had a milkround to cover and I’ve gone out after tea (evening meal), done half of the round, got home for midnight, up at 5 to milk the cows and then finish the round afterwards, all the while we had customers ringing to say they haven’t had their delivery yet.” His own summary of those years is simple: “I was constantly plate spinning.”

The community felt it. “Thank you, Cooil’s Dairy Ltd for your service over the years in rain, hail, and sun — but usually in the dark,” the Ballasalla Village community page posted. “It’s the end of an era”. 

The Collapse of the Middle

Cooil’s closure fits a pattern that’s been tightening for decades. In the early 1970s, an estimated 99% of UK milk was delivered to doorsteps, according to Andrew Ward’s No Milk Today. By the late 2010s — before the pandemic temporarily reversed the trend — that share had fallen to roughly 3%. But the closures aren’t spread evenly. They concentrate in a specific zone. 

At the top end, scale operators grow. McQueens Dairies in Scotland expanded well before COVID, with turnover climbing 30% in the year before their 2019 facility announcement, per BBC Scotland. By early 2021, they’d opened 11 distribution depots across Scotland and northern England and recruited more than 200 new staff in 12 months. 

At the bottom end, micro‑operators survive by stripping overhead to nearly zero. Gareth Baird, a young farmer in Carrickfergus, Northern Ireland, launched his doorstep round in July 2020, targeting 30 bottles on his first night — he delivered 120. No employees. No bottling line. Minimal fixed cost. 

The operators disappearing are the ones in between. Family dairies milking 80 to 200 cows, running their own on‑farm processing, employing a small team, and serving a few hundred to a couple thousand retail customers while sending the bulk of their milk to a cooperative. Cooil’s — 120 to 130 cows, five people, 850 houses at the end (just over 1,000 at the COVID peak), 80% of milk to the Creamery — was textbook middle‑zone. And the mechanics that made it unsustainable aren’t unique to the Isle of Man. 

Why the Math Stopped Working

If you’re running on‑farm processing, the number that matters most isn’t your customer count. It’s your retail‑to‑wholesale ratio — the share of your total milk output that actually flows through your bottling plant.

At the time of closure, Cooil’s sent approximately 80% of its milk to the Isle of Man Creamery at the cooperative wholesale price, per Manx Radio and 3FM. But that ratio had been worsening. In his Manx Nostalgia interview, Juan described it as “approximately about three quarters” to the Creamery, explaining: “We milk more cows. We’ve got more sort of surplus if you like, and then the rest goes on to the doorstep”. Every cow they added sent more milk to wholesale because the doorstep rounds couldn’t absorb the growth. By the end, only about 20% flowed through the family’s own pasteurizer, bottler, and delivery rounds — and that 20% had to carry the entire fixed cost of processing equipment that costs roughly the same whether it handles a fifth of the herd’s output or all of it (20% utilization means each litre through the bottler carries 5× the fixed cost it would at full capacity). 

Here’s what that equipment costs to replace. At the micro end, a basic batch pasteurizer starts around $14,000 USD (Tessa Dairy Machinery), while a complete micro‑processing system runs roughly $18,000 USD (MicroDairy Designs). But Cooil’s wasn’t micro — they were making skimmed, semi‑skimmed, and whole milk plus cream for hundreds of doorsteps. The Northeast Dairy Business Innovation Center’s 2024 Processor Modernization grants show real costs at this level: awards ranged from $62,000 to $350,000 per facility for vat pasteurizers, rotary filler‑sealers, and packaging lines. 

Take a bottling line with, say, a 12‑year service life and a $150,000 replacement cost. That’s $12,500 set aside every year just to fund its own successor. Spread that across 850–1,000 customers, and it’s roughly $12–$15 per customer per year in depreciation alone — before energy, bottles, labor, fuel, trailer haul‑back, or vehicle maintenance. And every hour Juan spent nursing a bottling line past its service life was an hour not spent on herd genetics, forage quality, or transition cow management — the core drivers of the 80% of the milk that actually paid the bills. That’s opportunity cost, and it compounds quietly. 

Cooil’s faced an additional constraint that most mainland operators don’t. The processing plant and the cows were on different sites. Every litre destined for doorstep delivery was pumped into that DX bulk tank on an old grain trailer, hauled back for processing five days a week, then bottled and delivered. That’s haulage, handling, and risk you don’t see on the milk cheque — but you pay for it.

They also operated under a regulated retail price ceiling. The Isle of Man’s Milk Marketing Committee sets retail milk prices by government order. The price per pint rose to 90p in July 2025, the first increase in roughly two and a half years. When your costs are rising, and your price ceiling is externally fixed, the only lever you have is volume. On an island of around 84,000 people, volume has a hard ceiling, too. 

QUICK CHECK: Is Your Retail Channel Paying Its Way?

  1. What would it cost to pay yourself and your partner market rate for every hour you spend on the retail side? That number is your unpaid family labor subsidy. If the retail channel can’t cover it, you’re already eroding — you just can’t see it on the P&L.
  2. Divide your processing equipment’s total replacement cost by its remaining service life in years. Are you banking that amount annually? If not, you’re consuming the asset without replacing it.
  3. Call your cooperative and ask one question: “Could you absorb our full supply volume within 90 days?” If yes, your safety net exists. Knowing that changes how you evaluate everything else.

Juan’s answer to those checks is pretty clear in hindsight. “We were understaffed really, and we could not afford to have relief staff,” he wrote, “and that’s if you could find someone who would do a bit of everything.” In the end, it fell to him to cover whatever needed doing. “All these factors led to my heart not being in the job,” he admitted. The bad days — the midnight‑home, 5 a.m. milking, customers ringing because the milk wasn’t there yet — weren’t constant. But they were frequent enough that, by the end, he was “almost begrudging having to do a milkround” while a new building on the farm and six kids at home all needed him too.

“So the end of an era, saying that seems to be the most commonly used phrase,” he wrote. “Yes, it is in its own way, so many people loved our products and are going to miss them.” But the other line that matters is this one: “At the end of the day, it wasn’t a decision that happened overnight, and we have made it for what we feel is right for us going forward as a family.”

Same Island, Two Models, Opposite Outcomes

Carl Huxham runs Cronk Aalin Farm in Sulby on the Isle of Man

A useful comparison sits on the same island. Carl Huxham runs Cronk Aalin Farm in Sulby, milking 40 cows but routing 100% of his output through nine delivery rounds using electric vans. Nothing goes to the Creamery. He built the operation from scratch, starting in 2006, buying a second‑hand 16‑point Fullwood parlour and bulk tank for £8,000. Every piece of equipment was sized for the volume it actually serves. 

“Being on an island, our input costs are all quite high as everything incurs a shipping cost,” Huxham shared. He sells at the same regulated price ceiling that Cooil’s operated under. 

The difference isn’t geography or regulation. It’s ratio. At 100% retail, Huxham’s processing equipment is fully utilized by the revenue it generates. At Cooil’s 80/20 split, theirs couldn’t be. If you’re considering building a farm‑direct operation from scratch, Huxham’s model is the template. If you’re inheriting one that’s already split between retail and wholesale, Cooil’s is the cautionary math.

Four Paths Forward

Across the UK and North America, family dairies navigating the same pressure points are finding distinct paths:

The clean exit to cooperative supply. Cooil’s path: cease retail, route all milk through the cooperative. It eliminates processing and delivery costs entirely, preserves the farm, and can execute within 60 to 90 days. The trade‑off is permanent — you surrender the retail premium and the direct community relationship. 

The radical downscale. Baird’s model in Northern Ireland strips out every cost layer that burdened Cooil’s: no paid delivery staff, minimal equipment, a radius one person covers before breakfast. It doesn’t scale. And it depends entirely on one body holding up indefinitely. 

The channel swap to vending. Milk vending machines have become one of the fastest‑growing farm‑direct channels in UK dairy. A setup costs roughly £30,000, with margins of 60-80 pence per litre, according to The Bullvine’s July 2025 analysis. The model eliminates delivery cost by bringing the customer to you — but also eliminates the community welfare function. 

The community‑supported model. Stroud Micro Dairy in Gloucestershire operates as a cooperative, with customers subscribing to seasonal shares. Over 800 community shareholders own Tablehurst and Plaw Hatch Farms in Sussex. When customers pre‑pay for a season’s milk, you know in January what February looks like. The catch: board meetings, annual reports, and governance paperwork most dairy families didn’t sign up for. 

PathEntry CostKey Trade‑offBest Fit
Clean exit to cooperativeMinimalLose retail premium permanentlyRetail <30% of output; equipment aging
Radical downscale$5K–$15KNo growth; one person’s staminaYoung farmers; no employees
Vending~£30K / ~$38KLose doorstep relationshipFarms near roads with footfall
Community cooperativeVariableGovernance complexityPeri‑urban; engaged consumer base

What This Means for Your Operation

This isn’t abstract. Clark Farms Creamery in New York was processing roughly 25% of the farm’s milk — about 3,000 gallons a week — with the remaining 75% still being trucked off. As The Bullvine reported in January, the real premium was roughly $1.15–$2.15 per gallon in extra margin at the cost of 70–90 more hours a week on top of a full dairy workload. Clark shut the creamery down in January 2026 while keeping the cows milking — the same decision Cooil’s made, on the opposite side of the Atlantic. 

Strategic PathEntry Cost (USD/CAD)Key Trade-OffBest Fit
Clean Exit to CooperativeMinimal ($0-$5K transition costs)Lose retail premium permanentlyRetail <30% of output; equipment aging; no succession plan
Radical Downscale$5K-$15K (minimal equipment, no employees)No growth; limited by one person’s staminaYoung farmers with no family; high energy; willing to work 7 days/week
Vending Machine Model~$38K CAD / ~£30K GBPLose doorstep relationship and community roleFarms near high-traffic roads; peri-urban locations; strong local brand
Community CooperativeVariable ($10K-$50K legal/admin setup)Governance complexity; board meetings; reportingPeri-urban locations; engaged customer base willing to invest; strong local food movement

On Vancouver Island, Mark at Promise Valley Farm took the opposite approach — a small organic Guernsey herd with 100% A2A2 genetics, processing all milk on‑farm through a self‑serve dispensing machine. “Processing our own milk and making value‑added products has to be part of the conversation for future producers,” he shared. But notice: small herd, all milk through the store. Ratio, again. 

The advantage you have that Cooil’s didn’t: pricing freedom. No government committee sets your retail price. You can charge what the local market will bear—but only if you use it deliberately. If you’re pricing farm‑store milk just a dollar above the grocery store “to stay competitive,” you may be leaving the margin on the table that would fund your equipment reserves. And here’s the piece that ties to your breeding program: if your herd’s component profile — butterfat, protein — commands premiums through the cooperative, every litre you divert to flat‑rate retail bottles is leaving that premium on the table. If you’re operating under Canadian supply management, the ratio math shifts because your wholesale floor is higher — but the equipment depreciation math doesn’t.

The wholesale safety net itself is under pressure. AHDB warned in January 2026 that farmgate prices are “set to stay under pressure into mid‑2026” as oversupply squeezes values, with the spring flush likely to make the first half of the year particularly difficult. The UK average farmgate price for December 2025 came in at 40.29 pence per litre, down 6.1% from November and 13% below December 2024. Exiting retail into a weakening wholesale market is still a viable move — but the window where wholesale alone feels comfortable is narrower than it was six months ago. 

Confirm your cooperative or processor can absorb your full supply before you need them to. Cooil’s transition to the Creamery happened smoothly on February 1 because that relationship was already in place. Your safety net should exist long before the pasteurizer starts making noises it shouldn’t

The Technology Temptation (Don’t Wait for It)

If you’ve heard about Lely’s Orbiter — an automated on‑farm processor that pasteurizes, homogenizes, and bottles with minimal labor — you might be thinking automation could change this math. As of December 2025, five units were operational in the Netherlands and Belgium, with sales expanding into Germany. The Orbiter page is live on Lely’s North American website. But there’s no announced NA availability date, no published pricing, and no regulatory pathway confirmed. Lely’s own Astronaut A5 Next milking robot won’t reach the US and Canada until “after local validation in 2026”. The Orbiter is further back in the queue. If your bottling line is at year 9 of a 12‑year service life, you can’t afford to wait for technology that may not arrive at a price point that works for your herd. Cooil’s made the call at their convenience. That’s worth more than any piece of equipment. 

Key Takeaways

  • Track your retail‑to‑wholesale ratio, not just your customer count. When your direct retail channel handles less than 30% of total output, the processing infrastructure is almost certainly overbuilt for the volume. Cooil’s saw that ratio worsen as the herd grew, from about 75% wholesale to 80%. Clark Farms hit the same wall at 75%. 
  • Count your unpaid family labor as a real cost. Cooil’s ran the entire processing and delivery operation with five people, in addition to the farm work. Add six kids and a second site into that mix, and you can see why Juan described his life as “plate spinning.”
  • You need an equipment replacement plan, not just a repair budget. If your retail margin isn’t building a reserve to replace the bottler, the day it fails, the decision happens to you, not with you.
  • Surge demand will lie to you. Cooil’s peaked at “just over 1,000” houses during COVID, then settled back to about 850 as people returned to normal. AHDB and Kantar saw similar patterns nationally. Don’t invest based on the peak; invest based on the plateau. 
  • Genetics and components matter to this decision. Every litre you pull from a high‑component herd and sell at flat retail is a litre that doesn’t earn the cooperative’s butterfat and protein premiums. That’s genetics ROI you’re giving away.
  • Year 8 is your red‑flag year. If your processing equipment is past year 8 of a 12‑year life, you should already be working through your options: full retail, clean exit, downscale, vending, or cooperative model. Not when something breaks. Not when Lely announces an Orbiter for your market. Now.

The Bottom Line

Juan and Kirsty Hargraves closed their retail operation while the farm was still healthy, the staff could be thanked by name, and the community had time to say goodbye. “We would like to think that nobody was ever let down,” they wrote. From Leslie Cooil’s first delivery in the early 1950s, through Ian and Gary’s decades behind the wheel, to Juan and Kirsty’s final round on January 31, 2026 — the milk showed up before dawn, to hundreds of doorsteps, every week. That’s not a failure story. 

The question for your operation isn’t whether something like this could happen to you. It’s whether you’d recognize the signals at year 8 — not year 12.

If the weight of a decision like this is sitting on you — or on someone you know — the Farm Aid hotline (1‑800‑FARM‑AID) and the Canadian Ag Mental Health Alliance can help. You don’t have to sort it out alone.

Complete references and supporting documentation are available upon request by contacting the editorial team at editor@thebullvine.com.

Learn More

The Sunday Read Dairy Professionals Don’t Skip.

Every week, thousands of producers, breeders, and industry insiders open Bullvine Weekly for genetics insights, market shifts, and profit strategies they won’t find anywhere else. One email. Five minutes. Smarter decisions all week.

NewsSubscribe
First
Last
Consent

The 143-Hour Week at Clark Farms: The Real Math of On-Farm Creamery ROI and Your Time.

Clark Farms operated a creamery for 6 years, serving dozens of accounts. They just shut it down—and kept milking. Here’s the math.

Executive Summary: Clark Farms, a fifth‑generation dairy in Delhi, NY, ran an on‑farm creamery for six years with dozens of local accounts, then shut the plant down in January 2026 while keeping the cows milking. Their numbers show what you’re really trading with on‑farm creamery economics: you’re not going from $1.85 milk to $5.50 milk, you’re buying roughly $1.15–$2.15 per gallon in extra margin at the cost of 70–90 more hours a week in processing and delivery on top of a full dairy workload. Backed by USDA, Rabobank, Cornell Dyson, and PASA data, the article walks through how consolidation, cost gaps, and thin processing margins make “just add a creamery” a much riskier survival plan than it looks on paper. It puts Clark’s pause alongside operations like MOO‑ville, Ronnybrook, and Hudson Valley Fresh that do make processing pay by staffing it as a true second business or sharing plants and brands across multiple herds, instead of piling everything onto one family. You also see how legacy, family bandwidth, and herd genetics change the risk math: a project that steals time from components, repro, and succession can quietly cost you more than it earns. The piece finishes with a clear playbook for your own decision—map out real weekly hours, set hard limits on account numbers and delivery time, build an off‑ramp before you pour concrete, and answer one non‑negotiable question: who’s actually milking while you’re bottling the milk?

on-farm creamery economics

A fifth‑generation New York dairy built a creamery with dozens of accounts, then hit pause in 2026—showing exactly how far the real numbers of on‑farm creamery economics can pull away from the brochure version. 

If you’ve ever thought, “We should bottle our own milk,” this one’s for you. Clark Farms in Delhi, New York, did almost everything by the book—stainless, brand, accounts, community—and still chose to shut the creamery down while keeping the cows milking. That decision says a lot about survival, hours, and the actual premium left after processing and delivery. 

The 2024–2026 Reality: Why On‑Farm Processing Looks Like a Lifeline

Let’s start with your world, not the grant brochure.

USDA’s 2022 Census of Agriculture dairy highlights show that the number of U.S. farms with milk sales from cowsfell from 40,336 in 2017 to 24,470 in 2022—a 39% drop in five years. Over that same period, the value of milk sales climbed 44%, from $36.7 billion to $52.8 billion, while the national herd sat around 9.3 million cows. Fewer farms. Similar cow numbers. More milk money is stacked on a smaller group of operations. 

Rabobank analysis calculated that in 2022, farms with more than 1,000 cows produced about 67% of U.S. milk, up from 60% in 2017. Brownfield Ag News, quoting Rabobank’s Lucas Fuess, reported that farms milking more than 2,000 cows carried total costs around $23.06/cwt in 2022, roughly $10/cwt lower than typical costs on 100–199‑cow farms. In a market where Class III, Class IV, or all‑milk prices bounce in the high teens to low‑$20s, that cost gap is the line between breathing room and quietly wondering how much longer you can hang on. 

In Delaware County, New York—Clark’s backyard—county plans and extension work describe a long slide in dairy farm numbers since the 1990s, leaving only a fraction of the former herds still milking. You don’t need a chart if you live there. You see it in empty barns and fewer bulk tanks meeting you on the road. 

Against that backdrop, on‑farm processing looks like a lifeline. In recent months, New York’s all‑milk price has often sat in the high‑teens to low‑$20s per hundredweight; USDA pegged the New York all‑milk price at $21.40/cwt in November 2024, roughly $1.85 per gallon before hauling and pooling. Farmstead creamery case studies and Cornell‑linked reports show small processors selling branded fluid milk in the mid‑single‑digit dollars per gallon, several dollars above that effective blend value. 

On paper, you’re trading $1.80 milk for $5‑plus milk. When the feed bill is chewing up your cheque, that’s a tempting trade.

Clark Farms took that road. Their experience shows you where the math—and the calendar—start to bite back.

Clark Farms: A Real‑World Test of the Dream

Clark didn’t just bolt a filler onto a corner of the milkhouse. They built a serious on‑farm creamery.

The farm sits on about 630 acres outside Delhi, New York, anchored by a barn Peter Clark built in 1907. A 2022 feature in Scribner Hollow describes it as a fifth‑generation dairy, already milking cows on that hill for about 114 years. 

Kyle Clark picked up the processing bug while studying dairy business management at SUNY Morrisville. Scribner Hollow reports that he spent roughly four years digging into New York regulations, working through inspections, and hunting down used dairy‑grade stainless pasteurizers, tanks, and bottlers from all over the country before the creamery opened in 2020. That’s not a spur‑of‑the‑moment pivot. That’s a long, careful build. 

By early 2022, the creamery was processing **about 25% of the farm’s milk—roughly 3,000 gallons a week—**with the remaining 75% still leaving on a tanker. That milk became bottled whole milk, flavored milks like chocolate and coffee, plus cream and butter under the Clark Farms Creamery label. 

They didn’t just sell from the farm store:

  • Their products moved through small groceries, cafés, and farm stands across the Catskills—Delhi, Andes, Phoenicia, Woodstock, and more. 
  • Cafés like Prospect and Fellow poured Clark milk into lattes and gelato and told that story to their own customers. 
  • Locals knew the Clark name when they opened a fridge door in town.

Demand wasn’t the snag. The cows were milking. The creamery was moving product.

On January 28, 2026, Clark Farms posted on Facebook that “after careful consideration,” they would be closing their creamery operations, while continuing to run the dairy farm. They wrote that “it has been a great joy to be able to produce dairy products for the community over the last six years,” thanked customers for their support, and admitted that keeping both the farm and the creamery to their standards had become too much. WBNG‑TV shared the news as “Clark Farms in Delaware County is closing its creamery doors after years of service,” and the comments quickly filled with people grabbing the last pints and thanking the family. 

So the hard question isn’t “Why didn’t people support them?” It’s “What did the money and the hours really look like when they chose to stop?”

The Big Math: What You’re Actually Trading

Here’s where on‑farm creamery economics stops being a dream and turns into a decision.

The Revenue Side

Picture a herd in the Clark range: roughly 200 Holsteins in milk, in a system where cows can produce around 80 pounds per day. Once you factor in dry cows, heifers, and the fact that no week ever runs perfectly, you’re in the ballpark of 9,000–10,000 gallons of milk per week for a solid 200‑cow Holstein herd in New York. 

If every gallon went into the pool at an effective $1.85/gallon (using that $21.40/cwt November 2024 New York all‑milk price as a real example), you’re looking at roughly $16,650–$18,500 per week in milk cheques at that volume, before hauling and other deductions. 

Now drop in the 25% processing share Scribner Hollow documented. In 2022, Clark was bottling about 3,000 gallons a week and shipping the rest as bulk. Say you can wholesale those 3,000 gallons at $5.50/gallon, right in the middle of what small on‑farm fluid and flavored milks often fetch locally. 

On a typical week, the rough revenue picture looks like:

  • Bulk milk: 6,000–7,000 gallons × $1.85 ≈ $11,100–$12,950.
  • Creamery: 3,000 gallons × $5.50 = $16,500.

Total: about $27,600–$29,450/week, versus $16,650–$18,500/week if every gallon went bulk at $1.85. On gross, that 25% slice looks like $9,000–$12,800 in additional cash flowing through the business.

That’s the number you hear in most creamery seminars. Now we get honest about what’s left after costs.

The Cost Side

Cornell’s Dyson School looked at 27 value‑added dairy businesses in New York, Vermont, and Wisconsin and didn’t sugar‑coat it: value‑added processing “is not a panacea.” In that study, mean net income from processing was modest at best and often negative, and average returns per cwt of processed cow milk were about $90/cwt lower than full economic costs once you charged a fair wage for family labor and a return on investment. The top performers did well, but the average small plant wasn’t swimming in cash. 

Farmstead creamery case studies from PASA and Penn State show direct processing costs (excluding milk) for small fluid plants often run about $1.00–$1.50 per gallon, once you add electricity, hot water, CIP chemicals, packaging, labels, maintenance, and required testing. That’s just to get a gallon into a bottle safely. 

Distribution costs pile on. Those same case studies document delivery expenses—fuel, truck payments, insurance, repairs, and driver labor—adding another $0.50–$1.00 per gallon in many rural, small‑drop routes. 

Now the comparison on a processed gallon looks more like this:

  • Extra revenue over blend: $5.50 – $1.85 ≈ $3.65.
  • Less processing + delivery: roughly $1.50–$2.50 per gallon combined.

That leaves a real premium of about $1.15–$2.15 per gallon before you pay yourself or cover downtime. On 3,000 gallons, you’re talking roughly $3,450–$6,450 per week more than sending that milk down the driveway.

Still serious money. But the “$5.50 instead of $1.85” story has already shrunk by more than half once stainless, cardboard, and diesel get their cut.

The Part Nobody Prints in the Brochure

For the days you’re reading this in the tractor cab, here’s the premium at a glance:

The MetricThe “Brochure” DreamThe Clark‑Style Reality (Illustrative)
Gross revenue$5.50/gal (wholesale price)$5.50/gal
Base milk value(Often ignored)($1.85/gal)
Processing & delivery“Minimal”($1.50 – $2.50/gal)
Real premium$3.65/gal$1.15 – $2.15/gal
The “price” you pay“Being your own boss.”70–90 extra hours/week

You’re not really selling $5.50 milk. You’re selling about $1.50 of margin and a second full‑time job.

The Time Math: Buying Margin with Hours

Here’s where on‑farm creamery economics stops being just dollars and starts being bodies and weeks.

PASA/Penn State’s farmstead creamery report logs weekly labor for small plants that easily hits 40–60 hours of processing and packaging once volumes reach a few thousand gallons. Add full cleaning and sanitation—CIP cycles, scrubbing floors and drains—and you’re realistically looking at about 40–50 hours inside the plant on a 3,000‑gallon week. 

Distribution and admin sit on top of that. For farms serving a couple of dozen accounts, those case studies show another 30–45 hours per week spent on ordering, route planning, loading, driving, stocking shelves, talking with store managers, invoicing, and chasing cheques. 

Now remember what the dairy alone demands. Cornell’s long‑running dairy farm business work shows that a 200‑cow Holstein herd with crops, youngstock, maintenance, and paperwork can easily soak up 50–70 hours a week from your core people. You probably don’t need Cornell to tell you that—you feel it in your knees. 

Put it together:

  • Dairy: 50–70 hours. 
  • Creamery: 40–50 hours. 
  • Distribution/admin: 30–45 hours. 

You’ve put in yourself well over 100 hours of work every week just to keep both sides upright. On a lot of family places, it feels like a 130‑ or 140‑hour week spread across three or four people, even if nobody ever writes it down.

Work CategoryWeekly Hours (Conservative)Weekly Hours (Realistic)Notes
Dairy Operations5070Milking 2×, feeding, bedding, calves, breeding, maintenance, crop work
Creamery Processing4050Pasteurizing, bottling, labeling, batch records, quality testing
Plant Cleaning & SanitationIncluded aboveIncluded aboveCIP cycles, floors, drains—often 8–10 hrs/week on its own
Distribution & Delivery2035Route planning, loading, driving, unloading, stocking shelves
Admin & Sales1015Invoicing, ordering, customer calls, chasing payments
Emergency/Downtime510Equipment breakdowns, inspector visits, surprise runs
TOTAL WEEKLY HOURS125180Spread across 2–4 family members—still unsustainable
$ Premium per Extra Hour$27–$49/hr$36–$90/hrBased on $3,000–$6,000/week margin ÷ 70–90 processing hours

Now take that extra $3,000–$6,000/week and divide it by the 70–90 extra hours wrapped up in processing and delivery. You’re effectively buying that premium at roughly $35–$85 of extra work per hour. On a spreadsheet that can be edited with a pencil. In a family that already feels stretched, it’s a different conversation.

At some point, many family‑run creameries, Clark’s included, look at that trade‑off and decide it no longer fits their standards for product quality, family time, or herd care. 

When On‑Farm Processing Really Works

This isn’t “never build a creamery.” It’s “be honest about what kind of creamery actually works.”

When You Build It for Scale

MOO‑ville Creamery & Westvale‑View Dairy (Nashville, Michigan). Doug and Louisa Westendorp started milking about 50 cows in 1992. When their six kids—including twins and triplets—wanted to stay on the farm, the family opened MOO‑ville Creamery in 2005 rather than chasing a multi‑thousand‑cow expansion. Today, Westvale‑View milks around 240 Holsteins, averaging over 100 pounds per cow per day, and MOO‑ville processes about 18,000 gallons of milk per week—roughly six times Clark’s processed volume. They’ve grown to four retail locations and products in over 140 retail stores and 50 ice cream shops. Their chocolate ice cream won first at the North American Ice Cream Association Conference in 2021, and vanilla followed in 2022. Every kid has a defined lane—herd, crops, ice cream, retail, tours—so nobody’s trying to juggle dairy, plant, and distribution alone. 

Ronnybrook Farm Dairy (Pine Plains, New York). Ron and Rick Osofsky started bottling unhomogenized milk in glass bottles at their Hudson Valley farm in the early 1990s. Thirty years on, Ronnybrook employs about 50 staff, crops roughly 760 acres, and sends milk, yogurt, and butter from their herd to 13 New York City Greenmarkets plus supermarkets across the region. In 2023, Scenic Hudson and Columbia Land Conservancy permanently conserved the farm, calling Ronnybrook “a local icon.” They built glass bottles and direct distribution into the business model from day one and staffed accordingly. 

In both cases, processing isn’t a side hustle tacked onto the dairy. It’s a second, fully staffed business.

OperationGallons Processed/WeekProcessing/Retail Staff (Approx.)Herd SizeOutcome
Clark Farms3,0002–3 (family)~200 cowsPaused 2026
MOO-ville18,00015–20+~240 cowsThriving, 4 retail locations
Ronnybrook~20,000+~50Large herd + 760 acresRegional icon, 30+ years

When You Share the Load

Hudson Valley Fresh (Hudson Valley, NY). Instead of each farm building a plant, 10 family dairies pool their milk into a shared processing facility at Boice Brothers Dairy, a family‑run plant dating back to 1914. Member farms—Jersey, Holstein, Guernsey, Brown Swiss, Ayrshire—must meet tight quality standards: somatic cell counts under 200,000, and raw bacteria counts under 5,000. In return, Hudson Valley Fresh has historically paid a premium, as evidenced by $23/cwt vs. $16/cwt for the commodity, through a base price plus quarterly profit‑sharing. The brand is strong, the creamery is centralized, and no single farm has to own all the stainless and all the route headaches. 

The pattern is pretty clear:

  • Some farms make processing work by building enough scale and staffing to treat it as a true second business.
  • Others make it work by sharing the plant and brand so their own time stays focused on cows and crops.

Clark’s situation—running a full dairy and a full processing/distribution business with essentially the same core people—is where a lot of smaller creameries stall out. 

The Accidental Trucker: When Your Farm Becomes a Logistics Company

When you first picture an on‑farm creamery, you see stainless steel and glass bottles. When it actually starts to succeed, you see routes.

Wholesale food moves on trucks, not spreadsheets. Dairy plant and creamery closure coverage—like Hastings Creamery in Minnesota or Prairie Farms plant changes—regularly points to transportation, labor, and maintenance costs chewing up thin margins. Grocery distributors live on low‑teens gross margins and 1–3% net margins once trucks, fuel, drivers, insurance, and warehouses are paid. They make it work with dense routes and big drops at each stop. 

A single‑farm creamery starts with none of that. You’re hauling your own product, on your own dime, to customers who might order heavy one week and light the next.

Somewhere between “a few good accounts” and “we deliver to everyone,” there’s a line. On one side, you’re still basically a farm that brings your own milk. Cross it, and you’ve become a logistics company that happens to own cows.

PASA’s case studies suggest that once you slide into the 15–20 account range, route planning, cooler space, and delivery windows start dictating your week more than milking times. With products in multiple groceries, cafés, and farm stands across the Catskills, Clark was clearly operating on that logistics side of the line. 

There’s nothing wrong with that if it’s what you want. MOO‑ville runs multiple delivery trucks and has staff just for routes and retail. Ronnybrook built those NYC market runs into its identity from the start. But you want to choose to become a logistics business, not wake up in one by accident

Why Clark Could Hit Pause and Keep Milking

Here’s a part of Clark’s story that deserves as much airtime as the shutdown: they could step off the creamery treadmill and keep the dairy running.

Look at a few structural choices they made:

  • They never processed all their milk. In 2022, about 25% went through the plant, while the rest still shipped as bulk. When they shut the creamery, they had somewhere to send that milk. 
  • They leaned into used, movable stainless. Scribner Hollow describes Kyle sourcing used pasteurizers, tanks, and other equipment from all over, rather than pouring everything into custom, immovable installations. That kept sunk costs lower and resale or repurposing options open. 
  • They didn’t strap the entire farm to the plant’s fortunes. We don’t see their loan documents, but the fact that the dairy stayed standing tells you the creamery wasn’t financed in a way that automatically dragged land and cows down when they hit pause. 

When WBNG shared the closure, locals piled into the comments talking about buying the last pints of milk and cream cheese and thanking the family for years of product. Another nearby business posted, “Sad news from our friends over at Clark’s farms! We’re wishing them the best and hoping to hear they re open in the future!” It read more like a community send‑off than a failure. 

That’s what an off‑ramp looks like when you actually need it. You don’t build it because you’re aiming to quit. You build it because you respect your dairy enough not to let one project drag the whole place down if the numbers or the hours stop lining up.

Legacy Changes the Risk Math

None of that math happens in a vacuum. Legacy sits right in the middle of it.

If you’re first‑generation with no clear successor, a creamery can feel like one more business shot. If it doesn’t work, you sell what you can, pay who you can, and move on. Ugly, but simple.

On a five‑generation place like Clark’s, the stakes hit different.

That 1907 barn overlooking Delhi was built by Kyle’s great‑great‑grandfather. By the time Scribner Hollow profiled the farm in 2022, they were already 114 years into the family dairy story. When the creamery pause hit social media, locals weren’t piling on—they were saying “thank you” and wishing the family well. 

Academics call it socioemotional wealth—all the non‑financial value tied up in keeping your farm in the family, protecting your name, and handing something real to the next generation. Put simply: how sick you’d feel being the one who lost the place. 

If you’re in Kyle’s boots, you’re not just asking, “Does this creamery cash flow in 2025?” You’re also asking, “If this ever pulls the dairy down, am I the one who ends 100‑plus years of work on this hill?”

That question doesn’t show up on your cost‑of‑production sheet. But it does show up when you decide whether to grind through another year of 140‑hour weeks or step back and protect the core dairy. Clark chose to protect the dairy. A lot of Bullvine readers would, too. 

Don’t Forget the Herd: Genetics, Components, and Long‑Term Value

There’s another engine running through this story that’s easy to overlook when you’re staring at stainless: your herd.

If you’ve spent years breeding for Fat, Protein, fertility, and health, that’s not just hobby genetics. It’s a big piece of your risk and return.

  • Milk cheques. In component‑priced systems, FMMO and co‑op schedules have been paying strong money for butterfat and Protein. When butterfat values in the Northeast climb above $3.00/lb, a herd running just a few tenths higher in components than the blend can easily see $2–$4/cwt more than average herds shipping the same volume. That shows up whether your milk goes into your own bottle or someone else’s cheese. 
  • Replacement pressure. Better fertility, health, and livability mean fewer heifers needed to maintain herd size. Analysts talking about the “processing gap” have also flagged tight heifer supplies and higher replacement prices, especially as more calves go beef‑on‑dairy. That makes every genomic and mating decision more expensive to mess up. 
  • Reputation and options. Strong cow families with performance and type give you options—embryos, breeding stock, bulls in AI, or simply better conversations with lenders and partners who know you’ve built something with resale value. 

If a creamery forces you to rob time from fresh‑cow checks, repro, and data review, you’re not just putting this year’s processing margin at risk. You’re putting the next decade of herd progress at risk, too.

When Clark stepped away from the creamery while keeping the dairy, they didn’t just protect cash flow. They protected a herd and a genetic trajectory, they’ve built one generation at a time. If you’re serious about breeding and components, that needs to sit right beside any stainless quote you’re considering. 

What This Means for Your Operation

Here’s where you stop looking at Delaware County and start looking at your own kitchen table.

1. Start with the One Question You Can’t Dodge

Before you chase a grant or sign a creamery loan, ask this out loud:

Who will milk the cows while I’m bottling the milk?

If your answer is “We’ll just work harder” or “We’ll figure it out,” you don’t have a plan. You’ve got a hope.

A real answer sounds like:

“I’ll run the plant from 9 to 3, my parents stay on both milkings for now, we hire a part‑time milker for two evenings, and if someone’s sick, we cancel one processing day and roll that product into the next run.”

If you can’t write down that level of detail—with names, hours, and backups—you’re not ready to buy stainless. You’re ready to go back to the whiteboard.

FactorBulk Milk (No Processing)On-Farm Creamery (25% Processed)
Revenue per gallon$1.85 (blend/pool price)$1.85 (bulk) + $5.50 (processed, 25% of volume)
Net margin per processed gallonn/a$1.15–$2.15 (after processing + delivery costs)
Extra weekly margin (3,000 gal)$0$3,450–$6,450
Weekly labor (family)50–70 hours (dairy only)120–160 hours (dairy + plant + delivery)
Effective hourly rate for extra workn/a$35–$85/hour (before equipment ROI)
Capital investmentMinimal (routine dairy equipment)$150,000–$500,000+ (plant, truck, cold storage)
Flexibility to scale downHigh—adjust herd size, cull strategicallyLow—fixed plant overhead, route commitments, debt service
Risk to core dairyLow—focus stays on cows and cropsHigh—time/capital diverted; plant failure can drag dairy down
Exit options if stressedSteady—co-op/processor always needs milkLimited—selling used stainless at discount, breaking customer commitments
Best fit for…Farms prioritizing herd genetics, land base, simplicity, or nearing successionFarms with committed next-gen, dedicated staff, dense local market, long runway

2. Put Your Week on Paper

Do this with the people who’ll actually be working it.

  • List every dairy job you do now with honest hours: milking, scraping, feed mixing, bedding, calves, breeding decisions, cropping, breakdowns, and paperwork.
  • List every processing job you’re adding: receiving, batching, pasteurizing, cooling, bottling, labeling, stacking, cleaning, swabbing, record‑keeping, and inspector visits. 
  • List every distribution job: loading, driving, unloading, stocking, talking with buyers, invoicing, chasing cheques. 

Add up those hours against the people you actually have—not the extra hire you “hope” will show up. If your plan needs any core person over 60–65 hours/week beyond a short start‑up push, be honest: you’re designing a burnout schedule, not a sustainable business. 

3. Decide When You’re Willing to Become a Logistics Business

Draw this line before you start adding accounts.

Write down:

  • The maximum number of wholesale accounts you’ll serve before you cost out a dedicated driver or route person—maybe that’s 12, maybe 18, but write a number.
  • The minimum drop size that makes a stop worth it—how many cases have to come off the truck to justify the fuel and time.
  • Which types of accounts you’re willing to walk away from if your route starts looking like spaghetti and you’re spending more time behind a café than behind a cow.

If you don’t set those rules now, your route will quietly run you.

4. Build the Off‑Ramp While You’re Pouring Concrete

If you go ahead, don’t wait until you’re exhausted to think about exit options.

  • Keep your co‑op or processor relationship alive, even if you’re shipping less. You want a place to send milk if you need to dial the creamery back.
  • Work with your accountant and lender, so the creamery debt doesn’t automatically drag land and cows into the fire if the plant has a bad year. 
  • Favor used or modular stainless steel, you could realistically sell if you change course. 
  • Set “stop rules” now: for example, “If after three years the creamery isn’t generating at least $X/month in net cash and our average weekly hours are still above Y, we pause and reassess.” Adjust X and Y to your reality—but don’t ignore them later.

5. Check Processing Against Your Herd Strategy

Your creamery plan shouldn’t be in a separate binder from your breeding plan. It should sit on top of it.

Ask yourself:

  • “Does this creamery help us capture more value from the components and health traits we’ve been selecting for, or does it risk pulling attention away from managing them?”
  • “If cash gets tight, do we cut repro and replacement investment first, or do we slow down creamery expansion and marketing?”

If the honest answer is “we’d sacrifice herd investment to keep the plant going,” you’re trading long‑term herd value for short‑term plant cashflow. That might be the right call in a specific situation. Just make sure it’s a conscious choice, not something you stumble into.

Key Takeaways

  • The on‑farm creamery premium is real—but it isn’t free. Processing a 25% slice of your milk might add several thousand dollars a week in margin, but you usually buy it with 70–90 extra hours of processing and delivery piled on top of a full dairy week. 
  • Once you’re past roughly 15–20 accounts, you’re running a logistics business. At that point, route density, drop size, and store delivery windows matter as much as butterfat tests. 
  • The farms that make processing work long‑term add people, not just stainless. MOO‑ville, Ronnybrook, and Hudson Valley Fresh all spread the load across family, staff, or co‑ops instead of forcing one family to carry everything. 
  • A well‑built off‑ramp is a form of insurance. Clark Farms could pause the creamery and keep milking because they never welded the farm’s survival to the plant’s success. 
  • Your herd and land are still the backbone. A strong Holstein herd on owned ground will outlast any value‑added project that only works if you live a 143‑hour week. 
  • The right time to discover your breaking point is on paper—not at 11 p.m. in the plant. If your creamery plan only works with everyone running at 110% forever, it doesn’t actually work. 
  • The first real due‑diligence question is simple. “Who will milk the cows while I’m bottling the milk?” If you can’t answer that clearly, you’ve already learned something important from Clark Farms’ experience. 

The Bottom Line

Clark Farms didn’t fail at diversification. They ran a full‑scale, real‑world trial of on‑farm processing under today’s economics, then listened when the premium and the hours stopped lining up. They made the hard call to protect the dairy that’s been there since 1907, honour their family’s work, and leave the door open for whatever comes next. 

If you’re weighing a creamery of your own—or you’re already living a week that feels a little too close to 143 hours—the real question isn’t just “Can this make money?” It’s “What does this do to our time, our cows, and our ability to hand something solid to whoever comes next?”

That’s the answer you want in hand before the stainless rolls into your yard.

Complete references and supporting documentation are available upon request by contacting the editorial team at editor@thebullvine.com.

Learn More

The Sunday Read Dairy Professionals Don’t Skip.

Every week, thousands of producers, breeders, and industry insiders open Bullvine Weekly for genetics insights, market shifts, and profit strategies they won’t find anywhere else. One email. Five minutes. Smarter decisions all week.

NewsSubscribe
First
Last
Consent
Send this to a friend