Like many of you, I was shocked to learn Brooklyn-based Farmigo was shutting down its online farmers market last week. In a blog post, CEO Benzi Ronen said the company could no longer continue its delivery service sustainably, so its engineers would be returning to Farmigo’s roots and focusing on its community supported agriculture (CSA) software platform.

It was surprising because the abrupt announcement came less than a year after Farmigo had raised $16M ($26 million in total). I had also just heard co-founder Benzi Ronen confidently claim that Farmigo’s software and community-focused distribution model would replace supermarkets this past May at the Future Food Tech Conference. And Ronen just wrote a piece for Food+Tech Connect, which highlighted many of the successes the company was having.

What happened? Earlier this week, Ronen and angel investor Ali Partovi, an early investor in Farmigo, spoke candidly with me about why Farmigo shut down its online farmers market, lessons learned and what’s next.

Is The Farm-to-Consumer Food Delivery Model Inherently Flawed?

Following Good Eggs’, another leading online farmers market, pull back of its operations in three cities last summer, my big question is: Is the farm-to-consumer delivery model inherently flawed, or have we just not found the right model yet?

Partovi believes it’s the latter. “Silicon Valley is sometimes guilty of hubris, thinking we can disrupt an industry overnight and do it better,” he says. “Sometimes we fail, but we keep trying and eventually we do succeed. I do believe the food industry is going to be made much more efficient, including farm-to-fork food delivery. I just think there’s a pull back happening, but I do think it’s not going to happen in the long run.”

Similarly, Ronen is convinced someone is going to upend the industry. In fact, Farmigo was actually nearing break-even and was profitable on a gross margin basis, according to Ronen. The decision to shutter the service was not for lack of market or feasibility, but rather because he and co-founder Yossi Pik felt that as software founders they were not best positioned to solve the increasingly complicated logistics required to meet evolving consumer delivery needs.

“We think we’re great at building software to solve problems, but we don’t think we’re great at building logistics to solve problems,” explains Ronen. “When you look at what the consumer is starting to want and where the environment is taking consumers, as far as how they want to consume better food and the skills you need in order to deliver on that, we started to see that we are no longer overlapping with where we needed to be.”

Consumer Shopping Needs Are Changing Fast

As companies like Amazon, Instacart and Fresh Direct begin offering 1-2 hour on-demand delivery, eaters are increasingly coming to expect this as the norm. This shift in is something that would have been difficult for Farmigo to predict even a year ago, says Ronen.

Therein lay Farmigo’s major challenge: its just-in-time, pick-up model was fundamentally at odds with on-demand or even next-day delivery.

With Farmigo’s community-centered food distribution model, people would place an order online, which farmers would then harvest and ship to Farmigo’s warehouse. Farmigo would then distribute the products to community organizers who would manage pick-up locations for local residents. By not holding inventory and removing last mile delivery to individual homes, Farmigo was able reduce costs and simplify logistics, but this approach also meant it took a couple of days for you to get your order.

The infrastructure Farmigo would need to support on-demand or even next day delivery is dramatically different than the system it had been working to perfect for the last 6 years.

After raising $26 million and building a community of employees, producers and customers that depend on you, its a difficult decision to walk away from your business, but Ronen and Pik knew this was not the right path forward for them.

Sustainable Food is Hard to Scale, No Matter How Much Capital You Have

Most venture capitalists are looking for high growth companies, so when you bring on capital your inclination is to grow fast and scale. As I wrote about when Good Eggs pulled back its operations, while it may be easy to scale software, many founders and investors don’t realize just how difficult it is to scale the distribution of perishable goods and brick and mortar infrastructure.

In some ways, Partovi thinks Farmigo and other companies that have had to pull-back operations did not have much of a choice. “There’s a combination of feeling seduced by the euphoria of other people doing more aggressive things,” he explains, “but also a legitimate eagerness to change and improve the food system faster.” There is also competitive pressure. If you’re not as aggressive as your competitors,  you risk being cannibalized and becoming irrelevant.

Partovi says he has always been impressed with how thoughtful Farmigo has been with the money they’ve taken. “When they did begin doing more on the consumer food delivery side it wasn’t nearly as aggressive a roll out as it could have been, “ he says. “It was relatively measured and careful, and I think they are in a better position to land on their feet.” He’s confident Farmigo will be just fine, because it still has its CSA software business, which has 300 to 400 customers, depending upon the season, and revenue.

What’s Next for Farmigo?

For now, Farmigo’s engineering team will focus on what they do best: software. Leveraging the deep understanding of farmers needs they’ve cultivated over the years, they hope to build a platform that enables farmers to best utilize the growing number of sales channels available to them.

What This Means for Investors & Startups

Farmigo’s shuttering of its delivery service is sure to make investors more cautious, but it shouldn’t deter them from investing in this space. And just because it may be more difficult to raise capital, it shouldn’t deter startups from launching new companies in this space. Savvy investors and entrepreneurs will recognize that grocery is a huge market with a massive opportunity to boost farmer’s profits and make it easier for them to get their products to eaters.

BUT this is a technology-enabled logistics business that is not for the faint of heart. Let me repeat: it’s a logistics business that is enabled by technology, not the other way around. Both Farmigo and Good Eggs took a software first approach.

Below is my full interview with Ronen, which has been edited slightly for clarity.

We will be digging deeper into the e-commerce space over the next couple of weeks, so sign up here to get notified when we publish more interviews and here if you’re interested in being featured.

____________________________________

Danielle Gould:  What happened with Farmigo?

Benzi Ronen:  We were at a juncture where we were about to raise another round, and these are the junctures where, as an entrepreneur, you have to ask yourself: Is this what I want to commit to for the next three to five years of my life? Is this the direction I believe is going to get the most leverage out of my time personally as a founder and what I could be doing? Is this going to be worthy of the amount of money that we’re raising from investors and able to give them a return on investment? Is it what the company can best do given the space and the environment?

We decided that continuing down the path we were on was not right for us. There’s definitely a passionate group of customers that want this kind of a service, but when we thought about what’s going to make the experience better for consumers, the answers we came up with were not improvements in the software and technology, they were improvements in the logistics platform. When I say logistics platform, I mean hardcore logistics, like how do we get the food to you faster? How do we reduce the order window from three days, since we had a just in time model where everything was pre-bought and farmers would harvest and deliver your orders to our warehouse, so when you place your order we give it to the farmers they would harvest it they bring it to the warehouse? How do we give you multiple orders a week? How do we enable home delivery from the pickup site to your home?

Basically, the product was the food, its quality and its transparency to farmers, and all the enhancements we saw in the years to come were around how to make the convenience of getting that food better and how to bring down prices. The answers [to how we would do this] were around getting more savvy at the logistics, while simultaneously continuing to improve our margins, which were also primarily driven by getting better at logistics, and then doing it well enough so we could expand into multiple regions, which is also how can we replicate those logistics. As software founders, we realized this wasn’t playing into our core competency, and so it was difficult for us to legitimize our expertise as being the best positioned to continue down our current roadmap.

DG:  When you launched the company, you must have realized this was a software and logistics business. How did your investors respond to your decision?

BR:  It was. I always called it a three legged stool: part software, part logistics and part this organizer model we had. The organizer model was Farmigo’s unique perspective on this industry. [We didn’t just want] to do home delivery like everybody else. What we had learned from the community supported agriculture (CSA) model was that pickup sites can be a very efficient way to distribute food and that finding someone in the community to take on this organizer role was leaning into a distributed model [that also gets] the consumer to play an active role.

Those are the three legs of the company, and we felt they were pretty evenly balanced. Software was enabling the customer to order, the organizer to manage their pickup sites and the warehouses to operate more efficiently. Obviously, we also had the logistics and our ability to recruit organizers, train them, launch them, manage a community of organizers. We felt it was pretty evenly balanced. And it was to date, but when we looked going forward, we really felt that most of the innovation would have to be on the logistics front and so the stool would be lopsided.

DG:  How did your investors respond to your decision?

BR:  It’s not like it happened in one day. We had a few different options of how to raise the next round and those options triggered a lot of different conversations. It’s not that all the investors looked at it the same way. Everybody had very different perspectives based on the funds they come from, the lifecycle of where the fund is, what they’re trying to get out of it, what kind of bets they want to make, how much more money they’re willing to put in. So it was a journey that we went [through] with them as well.

It’s not like I had figured all this out five months ago. You’re on the fundraising treadmill, and you’re running so quickly that you don’t have much time to stop and think. Victory is just closing a round of funding, so you can get the money and keep doing what you’ve been doing.

It took some time for me to finally say:  Wait a minute. I actually want to take a week or two with my co-founder before we close a round and not think about how I pitch and close the next round, but really do little bit more soul searching to figure out what we believe in and what we think we can pull off in a meaningful way. When I reached the decision and shared it with our investors, they were pretty much all understanding and supportive.

DG: That must have been really difficult to take a step back and reflect. How do you think you were able to do that?

BR:  I think when the funding is relatively easy – it’s never easy, but in all my previous rounds I had my pitch, the pitch never really changed and it’s a matter of kissing enough frogs until you find the right partner that you’re going to move forward with. Then you move forward with them and start executing the plan you just sold to them.

What happened in this case was we got an offer that made us pause and say: We just sold this plan, are we honestly ready to move forward with this plan? Is this the right plan? Are we sure about it? It was the type of investor we knew would take the plan very literally, and so that was question mark number one. So that made me take the plan back and revise it a bit, but I was still in the mindset of that I knew exactly where I was heading.

The second time that made me pause, [it made me think]: wait a minute before I put down another plan or proposal let me make sure that this resonates with my inner sense, and [I’m] not just [doing this] to win. As an entrepreneur, you get into this cycle where you just want to win. Losing is not an option. It’s very difficult to stop and say: I might not be on a winning path.

If you follow some of the conversations around knowing when to pivot and when not to pivot, it’s a really interesting debate. You can’t play out two different paths, so a lot of companies might be pivoting early, and they just didn’t have the wherewithal to play it out, to find out whether that would have been a successful path.

How do you know whether you’re pivoting too early? How do you know whether you’re pivoting too late? That maybe you’re just a stubborn entrepreneur that’s banging his head against the wall, and it’s never going to work. If you would just be a little bit more open minded and shift direction, then you might be on a much better path.

Jim Collins, author of Good To Great, has what he calls the hedgehog strategy. There’s three concentric circles: one circle is what are you passionate about, another circle is what are you best at and the other circle is the economic engine, the environment and what there’s demand for. Unless you’re working on something that meets all three, then something is wrong.

Am I passionate about the food system and fixing the food system? Do I love what I’ve been doing for the last eight years and Farmigo? The mission we set? Yes, I love it all.

Is there a need for what we’ve been working on? There is definitely a need for it – you can debate, I think, the magnitude of the need. How many customers are willing to pay a premium for a better product? There are certainly the die hard early adopters out there, and that’s who we were hitting. I think there’s another question which is: there’s no question people want better food if you can give it to them at the same prices, for sure, but that’s hard to do in the local food system today.

As Amazon starts to set a new standard, just like they did with free shipping in a day and now everybody has to get free shipping, and now it’s same day [food] delivery that they will do within two hours to your doorstep through a Prime Membership. Is that going to become the new standard? Maybe, and then people won’t want to compromise on that either. There’s the demand side of things, and that’s something we were dealing with and it’s not going to change overnight. But I think you need to get ready for [the demand side of things to change] and that’s part of where better logistics acumen comes into play.

Then it’s what are you great at? We think we’re great at building software to solve problems, but we don’t think we’re great at building logistics to solve problems. When you look at what the consumer is starting to want and where the environment is taking the consumers, as far as how they want to consume better food and the skills you need in order to deliver on that, we started to see that we are no longer overlapping with where we needed to be. And that’s a scary insight to come up with, especially when you’re knee deep into it. I don’t think it was an insight we could have come up with even a year ago.It’s just things were not heading in that direction.

We’ve built an incredible software platform that powers our warehouse – that’s what we’ve been building over the last year, as well as the supply chain that connects the farmers, consumers and organizers. That’s what we told our investors a year ago, and that’s what we built. But it was definitely scary. I mean it’s like holy shit.

DG:  From a logistics perspective, do you think the distributed organizer model made the logistics more challenging?

BR:  No, it simplified it for us. We had a huge competitive advantage on cost and logistics because we organize our customers into these pods. Think about Fresh Direct: if you and I live close to one another and you order your stuff to be delivered at 10 am and I order my stuff to be delivered at 11 am, we’re on completely different routes, even though from density perspective we’re neighbors. What we did is the organizer became a buffer so that we could deliver 20 orders to one location and then people could pick up at different times from that organizer. We even started to experiment with delivering out from the organizer to their home.

DG:  Can you share a number for how much of a competitive advantage it gave you? How much did it improve your margins?

BR:  It’s all comparative to each company. We were doing those benchmarks, but it’s going to be different for each one and it’s awfully different per city. To deliver in Manhattan, Fresh Direct actually has some advantages. You can almost see them doing that star in Manhattan where they’ll park the truck and then they’ll have runners go out from the truck, and so they’re actually doing the equivalent of a pick up location spot. But then you’ve got to Seattle and that model will never work, and that’s why Fresh Direct hasn’t been able to pull out of New York Metro.

DG: So that’s what worked with the model. What do you think didn’t work?

BR: I think anything that is logistics heavy takes time to figure out. We were getting much better at becoming profitable on a gross margin basis, but that takes work. It’s not like an investment in [pure] software, and then you start to see it payoff the more customers you have [as] you amortize your software investment across multiple customers.

It’s like every day you get just a little bit better packaging and the materials. You find another vendor and you get better at packing, so you can get more stuff [packed] into the materials you have.

Then you have labor and how do you give them the tools that they need in order to work effectively? How do you motivate them? How do you train them? What is the right balance between the hourly workers that were full timers for us and the contract labor we had to bring on during peak pack periods? How do you get better at your routing? We were outsourcing our deliveries to different vendors or independent drivers in each region, so how do you figure that out? Even the size of the vehicles changes.

I’ve just touched the tip of the iceberg. The level of complexity and iterations that you have, and then you mess up on one thing like throughput in the warehouse, and you just messed up all your routes. It’s a constant refinement, and it takes a certain type of person who’s really good at that and who can over time to continuously make improvements there. We have a wonderful VP of operations, best ever.

It’s not like I don’t think it’s possible. I thought it was possible. We were getting much better. What we were able to do even in the last half year is phenomenal. We were getting to a break even point. On a gross margin basis we were profitable. On a fully loaded gross margin basis – when you take into account customer support, sales, salaries, everything that is variable that’s needed company-wide to produce an order – we were kind of just getting to break-even, which was monumental. But we thought forward how we can make it even better, it just again comes back to: Is that what I am great at? No.

DG: What will it take to tackle these logistics, as far as talent and capital?

BR:  I think it depends what your appetite is. If you look at Winder Farms  and Farm Fresh To You, they’ve been doing this for about 20 years, actually Winder raised some money. If you started a small logistics operation and iterate around that, that’s one way to do it. If you want to move faster, a la Good Eggs , then you need to make some bets and you need to bring in an expert that you’re not breeding in-house, but you’re actually buying that expertise. There are different routes to get there. I don’t think there is any one way.

DG:  What about scale. Do you feel like you scaled Farmigo at the right time or was it premature?

BR:  We were a lot slower than, let’s say, Good Eggs. They immediately went into four markets. We were in New York forever (expanded to New Jersey, but that was out of the same warehouse), and then we went to San Francisco. Seattle we launched in October, because already investors were saying we’re not sure you know how to replicate the model. So did we grow too quickly? I don’t know.

We wanted to see if the organizer model could take root in other markets. That was a big proof point we needed to show investors in order to raise more money. So we had to prove that, and we couldn’t just stay in core markets, which were the Bay Area and New York. If it was a self-funded business, would I have leaned over that way? No.

DG:  What will happen to all of the producers who were selling through Farmigo?

BR:  Summer has historically been Farmigo’s seasonal low. Our school sites shut down and folks go on vacation. So our producers knew not to count on us for the Summer. Most of them make up for it via farmers markets and their CSA.

We are currently looking for partners that can service our customers once the Summer ends. We have also sent all our customers in each region a list of the producers they love so that they can purchase directly from them – Seattle-Tacoma, Greater New York and Bay Areas.

DG:  You scaled back the team significantly. Who remains on the team?

BR:  We had to lay off all the folks that were related to our community business, the consumer facing business, but anyone who was servicing the CSA business and the engineering team we kept on. I’m not going to give numbers, but it was the majority of our employees.

DG:  What does the next iteration of the CSA platform look like? 

BR:  We’re looking at it broader than a CSA platform. Right now we’re servicing CSAs very effectively, but we now we have experience working with a lot broader array of farms than just CSAs. We were working with all kinds of farms, fishermen, bakers and food artisans that wanted to get their food to market. Some of them were CSAs, and some of them were even using our software, but the majority weren’t. We feel that we now know better what software tools we can give all of them, so they can sell direct to consumer if they want, to food hubs if they want, to wholesalers and distributors, to retailers, to farm stands, farmer’s market. As a business person you should have a very diverse channel and not put all your eggs in one basket. I think you need software tools to figure out how to best utilize each one of those channels, and I think we’re in a pretty good position, as far as what we understand now for their needs, to build that kind of software.

DG:  How many how many customers are currently using your CSA Platform?

BR:  It’s seasonal, so it’s between 300 and 400, depending upon the season.

DG:  What do you know now that you wish you had known when you started Farmigo?

BR:  Farmigo started as a software company. We were building software to power CSAs. We had built a nice business that we never stopped operating and that’s the business that we’re throwing all of our development resources at. The consumer side was so appealing as an entrepreneur and for investors, because a lot of big exits happen when you’re able to do a large consumer play. At the time, when I was doing the CSA software, I didn’t see the CSA industry growing fast enough, so that that could become an interesting enough play for investors.

What I see today is that the local food system has been growing. There’s a lot more passion both from the farm side, business side (i.e. food hubs) and the consumer side. I mean you’re seeing Wal-Mart getting on the organic side, Whole Foods certainly with their sections and restaurants that want to be sourcing from farms. The industry as a whole has grown significantly, but it’s still incredibly inefficient to source this stuff — and those are the kinds of problems I know we can fix. As a software guy, as a guy who’s been building business to business software applications with SAP that connects businesses and allows them to trade more effectively together, to trust one another and to have transparency, to me that’s a massive opportunity. It’s one that I didn’t think was big enough in 2012 and 2013, and today I think it is.

DG:  What’s your advice to entrepreneurs who are entering this space?

BR:  Really perfect the model in one region, in all respects. Get the software right, get the experience right, get the logistics right, understand how your unit economic engine is going to work, so that you can get profitable. Really have an understanding of all of this before you start expanding.

There is [also the]  question of when is the right time to raise money, because when you raise money there is an expectation that you are going to spend money. So you’ve got to find the right balance. You certainly don’t want entrepreneurs who are starving, and you need some money in order to experiment and try and fail at things, so I think that would be the tightrope I would work on with the entrepreneurs; Experiment and figure things out locally, raise some money in order to get that experimentation and work through it quickly. Otherwise it’s going to take you a very long time, and then when you start to really have an understanding of what that blue print is about, then you can start pushing more aggressively outward.