Work Method Foundry Field Notes About Edge Brief

The Foundry · a Modven thesis

One problem, paid twice.

Every operating problem in your company holds two kinds of value. The first is the profit you get back when the problem is solved, which is priced at your own multiple and holds up when you sell. The second is the solution itself. If your problem turns out to be common across your industry, the fix can become an independent company, and you hold founding equity in it.

Read the thesis →

The thesis

Two assets wearing one disguise.

Most owners collect neither. They live with the problem, or they buy generic software that was built for a different company’s version of it. The Foundry is my process for collecting both, and the order matters: the certain payoff comes first, and the equity is what a well-chosen problem can earn on top of it.

The reason this belongs to the mid-market is an inversion that large companies cannot escape. A founding stake worth two million dollars is a rounding error to a billion-dollar enterprise, which is why the big venture builders only pursue nine-figure ideas. To a twenty-million-dollar company, the same stake is ten percent of enterprise value, acquired from a problem the owner was going to pay to fix anyway. In other words, the smaller company gets the outsized benefit from the very same stake. That is a big part of why nobody has offered this below the Fortune 500 before now: the wins are too small for their cost structures, and exactly the right size for yours.

One thing I will never do is pitch you startup paper as enterprise value. When a buyer looks at a minority stake in a young company, standard valuation practice discounts it by roughly 44 percent before anyone even asks whether the business is any good, and the accountants usually strip it out of the purchase price entirely. So I count the recovered profit as the return, I treat the equity as an option, and I never confuse the two.

The two harvests

First, I find the profit. Then the profit can become a company.

This is the moment where clients usually stop me: “So the system that fixes my margin problem… could also become a company I hold founding equity in?” Yes. That is the entire thesis, and everything else on this page exists to make it safe to believe. One is a promise and one is an option, and I never price them the same way.

One build, carried the whole way. Steps one through three are the engagement, and step four is what a well-chosen problem earns on top.

Harvest 01 · EBITDA, the quantified promise

First, the profit comes back to you.

The build recovers profit that was leaking out of your operations, and that profit is worth what everything else in your company is worth: your own multiple. A buyer’s accountants will credit every dollar of it, because it lives inside the earnings they are paying for. This is the only number I put in the headline, and a client who stops here got full value. Drag the sliders to your own numbers and watch what the build does to the value of the business.

value today
$19.8M
after the build
$24.8M

+$5.0M of enterprise value (+25%)

assumes the build recovers profit equal to 3% of revenue, the worked example on this page · value = ebitda × multiple · illustrative, not a quote

Harvest 02 · Equity, the same build carried forward

Then the fix can become a company.

Here is the part that changes how owners see their own problems. If yours is common across your industry, a generalized version of the solution becomes an independent company with a recruited founder and design partners from across your vertical. It is never built on your instance or your data, and you hold founding equity from day one. Here is the honest math. Suppose the new company reaches $2M of recurring revenue across twelve design partners by year three. At five times revenue, which is the plain multiple for niche business software rather than the fantasy one, the company is worth about $10M and your 20 percent is worth $2M. That stake is hard to sell and it is a minority position, but it is real, and it cost you nothing beyond a build you were funding anyway. I never count it as enterprise value until it converts to cash.

$2M
illustrative: 20% of a $10M company at year three · a bonus, never the promise

Only about one software startup in ten ever reaches $10M of annual revenue, and about half reach $1M. I put that number in front of you before you ask, because the whole model depends on respecting it. Most problems should stop at the build, and the screen exists to say so.

The deliverable

Every finding lands on one map.

Everything I find in your data gets plotted on two axes: how much profit the problem is costing you, and how common the same problem is across your industry. Where a finding lands tells us exactly what to do with it, and the map is yours to keep whatever happens next.

The Opportunity Matrix Vertical axis: EBITDA recoverable inside your business. Horizontal axis: prevalence of the same problem across your industry. Four quadrants: fix it quietly (high EBITDA, low prevalence); build it, then spin it out (high EBITDA, high prevalence); ignore it (low EBITDA, low prevalence); let someone else build it (low EBITDA, high prevalence). FIX IT QUIETLY HARVEST ONE ONLY BUILD IT, THEN SPIN IT OUT BOTH HARVESTS IGNORE IT LET SOMEONE ELSE BUILD IT EBITDA RECOVERABLE INSIDE YOUR BUSINESS PREVALENCE OF THE SAME PROBLEM ACROSS YOUR INDUSTRY LOW HIGH HIGH
Most of what I find lands on the left half of this map, and that is fine, because those fixes still pay for themselves. The Foundry exists for the findings that land in the upper right. The dots are illustrative findings from a single data room.
build it, then spin it out — high EBITDA, high prevalence fix it quietly — high EBITDA, low prevalence let someone else build it — low EBITDA, high prevalence ignore it — low EBITDA, low prevalence

the matrix is the deliverable you keep, regardless

How it runs

It runs in five steps, and the first one costs $500.

00

Edge Brief · Foundry Screen

Every engagement begins with the Edge Brief: a paid working session and a written brief, $500, credited in full if we build. It now includes the Foundry Screen, a one-page score of whether your problem is yours alone or your industry’s.

01

The data room

You deposit the operating data we request into a secure, access-logged environment, under an NDA that assigns all IP derived from your data to you. My agents analyze your operations against industry benchmarks, competitor postings, and supplier signals.

02

The opportunity matrix

Every finding is plotted by EBITDA recoverable and prevalence across your vertical. The matrix is the deliverable you keep regardless of what happens next.

03

The build complete on its own

I build the solution with the standard Modven machinery: dashboards, agents, algorithms, workflows. Cash-funded, and you own the system outright. The EBITDA lands here, and this step is complete on its own.

04

The second harvest

Only for builds in the upper-right quadrant. A generalized version is placed into an independent company with a recruited full-time founder and three to five design partners from across your vertical. You hold founding equity from day one.

The structure

Who owns what, and why.

This is who owns what when a new company is formed. The ranges follow venture-studio norms, and the logic comes from studying the spinouts that died because a parent company kept too much control.

PartyStakeWhy
Founder + option pool40–50%A recruited, full-time founder runs the company. Investors walk away when non-operating owners hold too much, so the venture must be fundable or it is worth nothing to anyone.
You, the client15–25%Founding common with a board observer seat and no veto. Your problem seeded the company; your control would suffocate it.
Modven15–25%A builder’s stake, earned on top of cash fees and never in place of them. Firms that traded fees for equity in the dot-com years stopped doing their best work, so I keep the two separate.
Design partners5–10%Warrants for the three to five industry peers who commit as first users.

The design partners are also the answer to the most reasonable objection in this business: that spinning out your solution arms your competitors. It cannot, for two reasons. The new company is built on generalized architecture plus data from peers who chose to participate, never on your instance or your data, and your named competitors can be excluded from its customer list for a defined period, in writing. More important, a product co-built with five committed companies across a vertical escapes the trap that kills internal tools turned startups: it is designed for an industry rather than extracted from one company. Your peers are not being armed with anything of yours. They are being recruited into a company you co-own, and their commitment is what makes your equity worth holding.

For operating partners

This model is at its strongest when you hold more than one company.

One portfolio company’s problem is rarely one portfolio company’s problem. A supply-chain gap in one $30M manufacturer usually exists in four others the firm already owns. Running the Foundry Screen across a portfolio does not just repeat the single-company offer several times; it changes the structure of the offer itself.

I know what this exercise used to cost, because I used to run it. In my venture-studio years we would surface a hundred problems across a portfolio, vote and rank them in rooms full of smart people, generate hundreds of candidate solutions, and emerge weeks later holding twenty to fifty ideas worth validating. It worked, and it was the most expensive imaginable way to learn what the portfolio already knew. That cost structure is gone. My agents run the same exercise continuously now — every operating problem in every company, surfaced, scored, and collided with live market data. The hundred-problem workshop became a standing process instead of a fund-cycle event.

Three things become true at portfolio scale.

01

The design partners are pre-assembled.

Sister companies replace cold recruiting, so a spinout launches with five committed users across the vertical instead of one. The single-customer trap that makes internal tools into weak startups dies on day one.

02

The failure math diversifies.

One venture is a lottery ticket. Eight screens producing two builds and one spinout is a portfolio strategy, which is the language a fund already thinks in. Base rates that are frightening at one company become planning assumptions at eight.

03

The firm gains a value-creation lever it can name to LPs.

Margin improvement now drives the majority of EBITDA growth in exited companies, and after four straight years of distributions below 15 percent of net asset value, funds are raising on the strength of their operating stories. “We convert portfolio operating problems into EBITDA plus a shared venture book” is a sentence no other mid-market sponsor can currently say.

And the exercise now produces something the workshops never could: an asset that outlives it. Every ranked problem and every killed solution carries priced reasoning, and that library compounds — across companies, across quarters, and into the next deal, where a fund that already holds its industry’s problem map walks into diligence knowing things the seller does not. The engine’s output is proprietary IP the firm owns, and until recently nobody could afford to build it.

The economics that keep the large venture builders working only with huge companies are the same economics that make a mid-market portfolio the natural home for this model. Each company’s build is paid for in cash and complete on its own. The option book is what the portfolio earns for holding companies whose problems rhyme.

Bring the one portfolio company whose problem you already suspect the others share. The Edge Brief is $500, credited, and the Foundry Screen will tell you whether the suspicion is right.

The Portfolio Screen · $25K, credited

Why now

The evidence, briefly.

10–12%
The annual profit growth a private-equity deal now needs, to earn the return that 5 percent used to earn in the 2010s. Operating improvement is the last lever left.
$3.8T
The combined value of 32,000 companies that private-equity firms own but have not been able to sell. The value now has to be built inside them.
51%
The share of profit growth in companies sold during 2025 that came from margin improvement, up from 21.5 percent before 2023.
95%
The share of generative-AI pilots that produce no profit impact at all. The demand is proven and the delivery is the bottleneck, and external builders succeed about twice as often as internal teams.
1 in 10
Software startups that ever reach $10M of annual revenue. I put this number on the page because the whole model depends on respecting it.
~44%
The discount a buyer applies to a minority stake in a young company before asking whether the business is any good. It is why I never count the option as enterprise value.
$12B
What Intuit paid for Mailchimp, a product built inside a small agency to solve its own clients’ problem. The pattern has famous winners: ServiceTitan, born in two family HVAC businesses, went public near $9B.

Questions

The honest answers.

“You’ll take my data and arm my competitors.”

No. Your data and your instance are yours, contractually and permanently, under an NDA that assigns the derived IP to you. Any spinout is built on generalized architecture plus data from design partners who chose to participate, and your named competitors can be excluded from its customer list for a defined period, in writing.

“The equity will turn out to be worthless paper.”

Often, yes, and I say so on this page. That is exactly why it is the second harvest and not the first. The EBITDA prices at your own multiple and survives diligence; the option costs you nothing beyond a build you were funding anyway. Compare that with any other way of acquiring a founding stake in a funded software company.

“If this works, why isn’t a big firm doing it for me?”

The economics. Venture building costs the large studios $500K to $1M per company, which is why every named builder serves clients above roughly $250M in revenue. The Foundry works below that line because your cash-funded build already paid for the asset and the spinout reuses it. One cost base feeds both engines, and an incumbent cannot follow without rebuilding its cost model.

“Who actually sells the new company’s product?”

The recruited founder and the design-partner network. Not you, and not Modven. Mid-market software sales cycles run six to eighteen months with five to ten stakeholders, and pretending a consultancy or a manufacturer can run that motion part-time is how spinouts die. The structure exists so that nobody has to pretend.

Start here

Start with an Edge Brief.

The Foundry begins the way every Modven engagement begins: a $500 working session and a written brief, credited in full if we build, with the Foundry Screen already inside it. One conversation tells you what your problem is worth once, and whether it might be worth something twice.

How the Edge Brief works