Why “Your Market is Too Small” is the Wrong Question

“Your market is too small.”

We’ve heard this concern in different forms over the past few months as we’ve worked through our strategic pivot—from serving Latin American general contractors across all disciplines to focusing exclusively on concrete structures in the US market.

Most investors supported the geographic shift. The US market, the infrastructure spending, the technology adoption—it all made sense.

But the specialization? That’s where the questions came.

The concern crystallized most clearly in an email from an investor who’d been following our progress:

“I’m supportive of the US move, but I’m struggling with the concrete-only focus. That’s an incredibly narrow wedge. Even if you execute perfectly, I’m having trouble seeing how this becomes venture-scale. How does ‘concrete estimation software’ become a billion-dollar platform?”

He wasn’t wrong to be skeptical. When you pivot from serving general contractors across every trade in an entire region to focusing on one structural system in one market, it looks like you’re thinking smaller, not bigger.

Here’s the thing, though: we’ve heard this objection before. In fact, we’ve heard it about every company that went on to define the digital economy.

They’re right. The market for “concrete estimation software” is tiny.

And just like Airbnb’s initial market was “air mattresses for conference-goers,” this tiny market isn’t the prize. It’s the key.

Because the question isn’t whether the current market is big enough. The question is whether the market is about to explode—and whether you can see it coming before everyone else.

Turns out, there’s a pattern to market explosions. Construction isn’t just showing the symptoms. It’s a perfect match.

And here’s what almost no one realizes: Starting a great ConTech company today is exactly like starting a great Fintech 15 years ago.

Let me show you why.

Act I: The Pattern

Finance 1999: What an Industry Looks Like Before It Explodes

In 1998, if you pitched venture capitalists on “software for financial services,” you’d get the same skeptical look we get today.

The total addressable market for banking software was measured in hundreds of millions, maybe low billions. Banks were digital deserts—most still processed checks by hand, moved money via fax, and spent less than 2% of revenue on technology.

The market was small because banks hadn’t digitized yet.

Then something changed. Not all at once—in waves.

Wave 1: The Infrastructure Layer (late 1990s - early 2000s)

Someone had to build the pipes first. Core banking systems. Payment rails. Basic digitization of the back office.

This created some solid businesses. But the TAM was still constrained by what banks were willing to spend on software.

Wave 2: The Wedge (mid-2000s)

This is where it gets interesting.

PayPal started as security software for Palm Pilots. A tiny, weird niche. Then they made a pivot that seemed small at the time but changed everything: they became the payment processor for eBay.

Suddenly, the TAM wasn’t “banking software.” It was “anyone who needs to move money online.”

New use cases created new markets.

Wave 3: The Explosion (2010-present)

In 2010, two brothers in Ireland launched Stripe. At the time, if you’d calculated their TAM based on “online payment processing,” you’d have concluded it was a crowded, commoditized market.

Three years later, in 2013, two engineers launched Plaid to solve a problem that seemed impossibly niche: aggregating bank account data via an API.

Investors asked the same question we hear today: “What’s your TAM?”

The fintech market has grown from roughly $200 billion in 2010 to a projected $1.5 trillion by 2030. That’s not because finance got bigger. It’s because what counted as finance got bigger.

Stripe, Plaid, Affirm, Brex, Mercury, Ramp, Nubank, N26, Monzo, Revolut, Wise, Chime, Coinbase…

The investors who said “too small” in 2010 missed every single one.

The investors who saw the pattern—who understood that PayPal wasn’t just Palm Pilot security, that Stripe wasn’t just payment processing, that Plaid wasn’t just bank data access—funded the companies that defined the digital economy.

Act II: The Symptoms

Construction 2025 = Finance 1999

Here’s the smoking gun.

Let me describe an industry, and you tell me which one I’m talking about:

The Industry:

Productivity has been stagnant or declining for 20 years
Spends less than 2% of revenue on technology
Relies primarily on paper, spreadsheets, phone calls, and faxes
Data is fragmented across dozens of disconnected systems
Workers spend hours every week manually copying data between tools
Projects routinely finish 20% late and 80% over budget
Facing a massive generational labor shortage
Young people don’t want to work there
Everyone says “this is just how it works” and assumes nothing will change

Finance in 1999? Or construction in 2025?

The answer is: both.

Let me show you the data:

Symptom	Finance 1999	Construction 2025
IT spend (% of revenue)	<2%	<1-2%
Productivity growth (20 years)	Stagnant/negative	+0.4% annually
Primary tools	Paper, fax, phone	Paper, spreadsheets, phone
Data systems per firm	Highly fragmented	11 different systems
Time on manual data entry	Hours per week	14 hours per week
Project overruns	Standard	80% over budget
Worker shortage	Looming	439,000 positions unfilled
Industry sentiment	”Can’t change"	"Can’t change”

The pattern isn’t subtle. Construction is showing every single symptom of a pre-explosion market.

But here’s what makes this a trillion-dollar opportunity, not just a productivity problem.

The $1.3 Trillion “Spend Gap”

After finance digitized, technology spending went from 2% to 6-12% of revenue. That 5-10 percentage point increase wasn’t just existing vendors getting more money. It was thousands of new companies building tools that didn’t exist before.

Construction is a $13 trillion industry currently spending less than 2% on technology.

If construction follows finance’s trajectory—and every indicator suggests it will—that spend gap represents $650 billion to $1.3 trillion in new software spending that doesn’t exist today.

This isn’t about capturing a slice of the current ConTech market. This is about the market itself expanding by 50-100x.

Act III: The Catalyst

Why Construction Stayed Stuck (And What Just Changed)

Okay, so if the pattern is so obvious, why hasn’t construction exploded yet? Why has it stayed stuck for 20 years while finance transformed?

The answer is brutally simple: finance had an easier problem to solve.

Finance’s data was locked in databases, but at least it was structured—rows, columns, numbers. The API was the key that unlocked it. Build an API layer like Plaid, and suddenly thousands of apps can access that data.

Construction has a fundamentally harder problem: its most valuable data is completely unstructured.

The information that determines whether a project makes money or loses millions is trapped in:

2D blueprints representing complex 3D structures
Architectural notation developed over decades
Hand-drawn markups on PDF plans
Building codes that vary by city, state, and country
Engineering standards that take years to learn
Email chains with critical change orders buried in paragraph 7

You can’t just “API” your way into a blueprint. You can’t write a script that reads an architect’s red-pen markup on a structural drawing.

This is why every previous wave of construction tech failed. They tried to force contractors to input data manually. “Please type all your quantities into our clean spreadsheet. Please log into our portal and update your schedule. Please photograph every delivery and tag it in our system.”

Contractors took one look and said: “That’s more work, not less. I’ll stick with my spreadsheet.”

The software died.

But Something Just Changed

Between late 2023 and early 2024, something extraordinary happened in artificial intelligence. A new category of models called “multimodal AI” emerged—systems that could process not just text, but images, documents, diagrams, and spatial relationships.

GPT-4 with vision. Gemini 2.0. Claude with computer vision.

For the first time in history, AI can:

✓ “Read” a construction blueprint like a human engineer reads it
✓ Understand spatial relationships in 2D drawings
✓ Extract structured data from unstructured documents
✓ Learn building codes and construction notation
✓ Interpret hand-drawn markups and dimensional callouts
✓ Get more accurate with feedback and training

This is the catalyst that didn’t exist two years ago. This is why previous attempts at construction automation failed—and why now is different.

The API unlocked finance. Multimodal AI unlocks construction.

The Labor Crisis: Why Adoption is Non-Discretionary

And here’s what makes the timing perfect: construction doesn’t have a choice anymore.

The industry is facing a labor crisis that makes technology adoption survival, not optimization:

439,000 open construction jobs in 2025
1 in 4 construction workers over age 55, approaching retirement
Major infrastructure projects delayed or canceled due to workforce shortages
A $40 billion semiconductor factory in Arizona stalled because TSMC couldn’t hire enough contractors

When the world’s most advanced chipmaker can’t build a fabrication plant because there aren’t enough construction workers, you know the industry has hit a wall.

AI isn’t a “nice to have” anymore. It’s how projects get built when you can’t hire enough people.

This is construction’s smartphone moment. The catalyst is here. The forcing function is here. The explosion is inevitable.

Act IV: The Opportunity

Why “Too Small” is the Wrong Question

This brings us back to that skeptical investor and his question about TAM.

The mistake he’s making—the mistake most investors make when evaluating construction tech—is measuring the current market instead of understanding the expansion pattern.

Think about what Plaid faced in 2013. Investors looked at the existing market and saw: “bank data aggregation for a handful of fintech startups.” Maybe a $50-100 million TAM.

They completely missed that Plaid wasn’t serving a market. Plaid was building the infrastructure layer that would enable thousands of new companies to exist.

Today, Plaid connects to 12,000+ financial institutions and powers apps used by hundreds of millions of people. The TAM didn’t grow because finance got bigger. The TAM grew because Plaid made it possible to build things that couldn’t exist before.

We’re building the same thing for construction.

Our “Plaid for Construction” Strategy

We’re not building estimation software. We’re building the data infrastructure layer that structures construction’s unstructured data.

It starts with estimation—specifically, the most complex type: reinforced concrete structures.

We chose this for three strategic reasons:

1. It’s high-stakes

Get a concrete estimate wrong, and you don’t just lose money. You halt a pour at 3 AM. You watch your schedule crumble. You eat six figures in wasted material or emergency procurement. The pain is acute, and contractors know it.

2. It’s complex

Reinforced concrete is one of the most data-dense, technically demanding construction workflows. It requires understanding:

Structural engineering principles
Local seismic codes and design requirements
Material properties and specifications
Construction sequencing and phasing
Dozens of interdependent quantities

If AI can master concrete estimation, it can master anything in construction.

3. It’s valuable

Here’s a misconception we hear a lot: “Concrete is only used in a small percentage of buildings.”

That’s true by count, but completely wrong by value. Concrete buildings may be 5% of structures built, but they represent 25-35% of total construction spending. A single 40-story concrete tower has the construction value of hundreds of single-family homes.

But as we stated in Our (Not-So-Secret) Master Plan, concrete estimation is the wedge, not the destination.

The Four Stages: From Wedge to Platform to Ecosystem

Stage 1: The Wedge (Now)

Use multimodal AI to read blueprints, generate BIM models and extract quantities with unprecedented speed and accuracy. Solve a single, high-pain problem better than humans can. Build trust. Collect data.

Stage 2: The Workflow

Once we have the foundational data structured, we connect it to the entire workflow. The estimate becomes the procurement order. The quantity data becomes the project schedule. We eliminate the 14 hours per week that project managers waste manually transferring data between systems.

Stage 3: The Platform

Scale beyond concrete to all trades—structural steel, MEP systems, architectural finishes, site work. Become the intelligence layer that structures data across the entire preconstruction process.

Stage 4: The Transformation

This is when it gets really interesting.

Once we have a structured data layer showing how billions of dollars of construction actually gets planned, estimated, and built across thousands of projects, we can start enabling capabilities that seem impossible today:

Generate construction-ready documents from architectural concepts
Predict schedule delays before they happen
Optimize designs for cost, schedule, and constructability in real-time
Enable true prefabrication at scale
Compress 6-month preconstruction cycles to weeks

This is when construction stops being artisanal and becomes industrial. When buildings start getting made more like cars: designed for manufacturing, assembled from standardized components, predictable in cost and schedule.

The “Dominant Player Gap”

Here’s the tell that construction is still pre-explosion:

Think about fintech. How many fintech unicorns can you name off the top of your head?

Stripe, Plaid, Nubank, N26, Monzo, Revolut, Wise, Chime, Coinbase, Affirm, Brex, Ramp, Mercury, Marqeta, Adyen…

I could keep going. You probably stopped counting at ten.

Now think about construction tech. Name the dominant players.

Autodesk, Procore. Maybe one more if you squint.

That’s… about it.

This is the smoking gun. When an industry has 2-3 dominant companies instead of 20+, when you struggle to name even five major players, you’re not in a mature market. You’re in the pre-explosion phase.

Finance in 2010 had a handful of established players. Today it has hundreds of unicorns and thousands of venture-backed companies.

Construction in 2025 has a handful of established players. In 2035…

The Prize: A $1.3 Trillion Market Gets Created

So let’s reframe the question the investor asked.

He wanted to know: “What’s your total addressable market?”

But that’s asking: “How big is the market that exists today?”

The better question is: “What happens when a $13 trillion industry moves from spending <2% on technology to spending 6-12%?”

The answer is: A $650 billion to $1.3 trillion market gets created.

Not captured. Created.

That’s the opportunity that investors miss when they anchor to the current TAM. They’re measuring the size of the old market—the one before the catalyst.

The companies that win aren’t the ones serving the existing market. They’re the ones building the infrastructure that enables an entirely new ecosystem to exist.

Why This Time is Different

Look, we know construction has tried to digitize before. We know there’s skepticism. We know contractors have been burned by software that promised transformation and delivered complexity.

But three things are different now:

1. The technology finally works

Multimodal AI can actually read a blueprint. Two years ago, this was science fiction. Today, it’s production-ready.

2. The forcing function is here

The labor shortage isn’t theoretical anymore. It’s existential. Contractors who don’t figure out how to do more with fewer people will go out of business.

3. The pattern is undeniable

We’ve seen this movie before. We know how it ends. The industries that looked impossible to digitize—the ones where everyone said “it’s too complex, too fragmented, too traditional”—those are exactly the ones where the transformation created the most value.

The Choice

Fifteen years ago, a small group of engineers and investors saw what everyone else missed: that finance was about to explode. They built Stripe, Plaid, Square, Coinbase. They defined the digital economy.

The investors who bet on “too small” fintech markets back then—who understood that PayPal wasn’t just Palm Pilot security, that Stripe wasn’t just payment processing, that Plaid wasn’t just bank data—funded the companies that shaped how the world moves money.

The investors who passed because “the market was too small” are still explaining why they missed Stripe at $5 million.

Starting a great ConTech company today is exactly like starting a great Fintech 15 years ago.

The players aren’t established yet. The data is locked away, waiting to be structured. The catalyst just arrived. The labor crisis is forcing adoption.

Every condition is aligned.

Today, construction is at the same inflection point.

So yes, our market is too small. Today.

But the question isn’t about today. The question is whether you can see what happens when a $13 trillion industry—the second-least digitized major sector in the global economy—finally catches up to the rest of the digital world.

The question is whether you understand that “too small” isn’t a problem to avoid. It’s a signal you’re early.

The question is whether you can recognize the pattern before everyone else.

We can. We’ve seen this movie before.

And we know exactly how it ends.

Today, construction is at the same inflection point.

If you’re an investor who recognizes the pattern before everyone else does.

If you’re an engineer who wants to build the infrastructure layer for a $13 trillion industry.

If you’re a construction leader who’s done watching technology fail and ready to help us get it right.

We’re building it. Join us.