Rebuild the world by backing the right founders.
This is the thesis Creci Ventures was built to test. We wrote it in 2023, before the fund had deployed a dollar, and published it unchanged. What follows is the argument, and then what happened when we put money behind it.
The largest thing the country builds, and the least of it runs on modern software.
Scale
Construction is 4.3% of US output. Real estate, rental and leasing adds another 13.6%. Close to eighteen cents of every dollar the country produces, and almost none of it runs on software built this decade.
Productivity
Construction labor productivity fell more than 30% between 1970 and 2020 while productivity in the rest of the economy doubled. The Richmond Fed puts construction at roughly a third of the decline in trend US GDP growth since the war, on the order of $1 trillion every five years. That is not a sector with a software gap. That is a sector that has gone backwards for fifty years while everything around it improved.
Capital and work, both repriced
Two changes reset the underwriting and neither has reversed. Where and how people work changed permanently, which repriced commercial space. And the cost of capital changed the math on every development pro forma. Founders who take real cost out of either one are selling into a need, not a preference.
Regulation is a records problem
Regulation is unavoidable friction in development, and underneath it is a records problem. Recording, tracking and aggregating what compliance demands is exactly what software does well, and exactly what most of this industry still does by hand.
Sources: US Bureau of Economic Analysis, value added by industry as a percentage of GDP, Q1 2026. Federal Reserve Bank of Richmond, “Five Decades of Decline: U.S. Construction Sector Productivity”, 2025.
We are realistic about how technology gets in.
Existing patterns of work
Development, construction and distribution run on decades of ingrained ways of getting things done. We do not believe an outsider talks the industry out of them. The founders who win are the ones who worked inside those paradigms and know which habit is load-bearing and which is only habit.
Fragmented markets
Real estate is geographically fragmented by nature, and so is everything built on top of it: financing, construction, design, distribution, management. That fragmentation is the hardest part of going to market here, and we underwrite for it rather than around it.
Three tenets, descending in scale.
Three tenets guided every allocation. They descend in scale: the market, then the business, then the person. They are reproduced here as written, before the fund deployed.
Market and technology movements
More of what people and businesses do is going to be modeled digitally, and more decisions will be made from that data. What Meta and Google did for marketing spend, thousands of companies will do for millions of other economic decisions.
We also believe the software layer keeps commoditizing. In an age of generated code the interface may dissolve entirely. What cannot be commoditized is the underlying data, and that is where the value settles.
None of which removes the need for judgment. If starting a company were a calculation, nobody would choose to do it.
Business properties
At the early stage we prefer companies already past their most violent course corrections. Surviving that volatility tells us two things: the founders have the stamina, and they have learned enough about their customer and their model to move faster with less left to discover.
The idea is Taleb’s antifragility. Systems tested early tend to last.
Founder characteristics
Most companies succeed because of determination, not because of the best idea or the smartest founder. And the most determined people are the ones being honest with themselves about whether they actually care about what they are building.
Honesty usually travels with humility, and the best operators we know have a real sense of their own limits. We take a humble warrior over a fearless leader every time.
A thesis is only worth anything if you publish how it held up.
Fund I deployed into two proptech companies, first close April 2024. Here is how each tenet has held up, graded in 2026. We are grading our own homework, which is worth what it is worth, but the thesis above was published first and has not been edited since.
The data centricity thesis
In 2023 the claim that the interface might dissolve entirely in an age of generated code was the most speculative thing on this page. It reads conservative now. The part we would strengthen is the corollary: if the interface is cheap, then distribution and proprietary data are close to the only defensible things left, and early stage diligence should weight them accordingly.
The post in-market inflection point
This is the one that failed, and it failed in an instructive way. One Fund I position cleared the inflection test on every measure we had. It had survived its product and go-to-market volatility. The position still did not work out the way we underwrote it.
The tenet is not wrong about what it measures. It is wrong about what it protects against. Surviving product volatility tells you nothing about counterparty risk, capital structure, or what happens to a company inside a transaction. We were measuring the business and the risk was not in the business. Any future version of this tenet has to be paired with diligence on the things that can take a good company away from you.
The honest founder mandate
At this size a fund is underwriting people, not a portfolio. There is no diversification to hide behind and no second fund to average into. Everything that went right traced back to a founder who kept going, and the judgment we would most want back is time spent on the person rather than the model.
And the finding that is not about proptech at all. Two positions is not a portfolio. A fund this size can test whether a thesis is coherent and whether its operator can source, diligence and work alongside a company. It cannot test whether the thesis produces returns, because the sample is too small for the answer to mean anything. That is the honest limit of a fund this size, and it is worth saying out loud rather than dressing two outcomes up as evidence.
If this thesis resonates with you, please reach out.
Fund I is deployed and we are not reviewing new founder pitches. The thinking above is still what we would back, and we are glad to talk to anyone working on it.