Henry Gordon Smith: The Man Behind the World’s Most Innovative Farms

Henry Gordon-Smith is the founder and CEO of Agritecture, a global advisory and technology firm that works on the planning and development of modern farms. Over the past decade, the company has advised agricultural projects across 45 countries, working with governments, developers, investors, and farm operators.

Gordon-Smith focuses on the earliest decisions in agriculture — what to grow, where to build, and how to structure a project — choices that shape outcomes long before results are visible. Across dozens of markets, he has seen the same pattern repeat: farms struggle when those foundational decisions lack structure, data, or scenario testing.

“The current way of planning farms of all kinds is really outdated.”

Improving how those decisions are made, and giving operators better tools at the earliest stages, has become the focus of Gordon-Smith’s work.

How His Perspective Was Formed

Raised in Hong Kong and Tokyo, Gordon-Smith grew up far from agriculture. His interest took shape during college, where he studied political science and resource conflict and became focused on food systems as a leverage point within global supply chains.

“I said the climate crisis is the biggest global challenge that I want to dedicate my career to.”

He began researching agriculture publicly, using writing to test where his interest and market demand aligned. Agriculture emerged as the clearest signal. In 2011, while still in college, he launched Agritecture as a research platform focused on how farms are planned and financed.

As demand for his work grew, Agritecture evolved from research into consulting, and eventually into an independent company with a global advisory team. That progression shaped Gordon-Smith’s focus on the early decisions that determine whether agricultural projects succeed or fail.

From Advisory to Infrastructure

Agritecture initially addressed this work through consulting, helping governments, developers, and operators assess feasibility, design concepts, and business cases across diverse markets. As the firm evolved, Agritecture launched a second phase of the business: Agritecture Designer, a planning platform that formalizes early-stage analysis. The platform allows users to evaluate locations, explore crop options, test market assumptions, and compare scenarios before engaging suppliers or financiers.

In one example, a vertical farm in Saudi Arabia used the platform to test scenarios, select suppliers, and secure financing. The planning outputs were used to obtain a $2 million government loan, significantly reducing the cost and time typically required for traditional feasibility studies.

“That’s a win for the operator, the suppliers, and the people financing the project,” Gordon-Smith said.

Vertical Farming, Explained

Vertical farming has become one of the most visible symbols of innovation in modern agriculture. By growing crops indoors under artificial light, often in stacked layers, these systems promise year-round production and proximity to urban markets.

Gordon-Smith says that appeal is often overstated. The assumption that these systems can be deployed anywhere, grow almost any crop, and deliver higher yields without significant tradeoffs does not hold up in practice. The economics are far more constrained.

Because artificial lighting replaces the sun, energy becomes a dominant cost, making vertical farms highly sensitive to electricity prices.

“If energy prices are too high, vertical farms can’t make any money,” Gordon-Smith said.

He points to Bermuda as an example he was asked to assess. Despite high food prices and limited land, electricity costs there can reach 36 to 40 cents per kilowatt-hour—levels that make even well-designed vertical farms difficult to justify.

Crop choice is another limiting factor. Vertical farming works best for crops where most of the biomass is edible, such as leafy greens.

“If I grow lettuce, most of the plant is edible,” Gordon-Smith said. “If I grow something like an avocado tree,

I’m paying energy to grow a lot of biomass I can’t eat.”

Yield gains are real, but they come with higher capital intensity and operational complexity. For Gordon-Smith, vertical farming is neither a cure-all nor a failure, but a tool that works well under the right conditions—and fails quickly without careful planning.

Where Agriculture Is Heading

Looking ahead, Gordon-Smith expects agriculture to become more data-driven, more regionally differentiated, and more disciplined about capital. Growth, in his view, will be uneven: steady in North America, expanding in Northern Europe, accelerating in Southeast Asia where costs are lower, and strategic—but cautious—in the Middle East.

More broadly, Gordon-Smith believes the next phase of agricultural adaptation will be defined by speed. As production conditions become less predictable and capital more selective, decisions that once unfolded over years will increasingly need to be made under compressed timelines.

“Things are going to get bad enough that we’re actually going to act,” Gordon-Smith said. “And when that happens, agriculture doesn’t move fast—but we’re going to have to move fast.”

When supply disruptions, price volatility, and resource constraints converge, he argues, there will be far less room for improvisation and far less tolerance for failure.

What Smith Is Building Toward

Gordon-Smith’s focus now is strengthening the systems that support decision-making across agriculture. Through Agritecture’s advisory work and planning platform, he aims to make agricultural complexity more legible—giving operators, investors, and policymakers clearer tools to evaluate options before capital is committed and infrastructure is locked in.

“I want to have the platform that allows people to access data, best practices, and models faster,” Smith said.

In a sector facing mounting operational and financial constraints, that kind of preparedness may prove to be one of agriculture’s most important advantages. The question, Gordon-Smith suggests, is no longer whether food systems will need to adapt, but whether the tools to support that adaptation will be in place when they are needed most.