Foreword
Almost everything built for a growing business is paid for before anyone knows whether it works.
This is so ordinary that it is invisible. You hire the firm, you pay the fee, the work arrives. Whether it moves anything is a separate question, asked later, by you, alone. The people you paid have already been paid.
This memorandum is about what happens when you refuse that arrangement — not by promising to be better, but by building a structure in which, when what we build does not work, we forfeit everything beyond our cost.
Everything here reduces to one sentence, and the whole apparatus exists to make that sentence checkable rather than charming:
We cannot win unless you win.
Not as a promise. Promises are what people offer when the structure doesn't force their hand. As architecture.
The Misalignment
Consider what you are actually buying when you buy growth work.
You are buying effort, priced as if it were outcome. The invoice arrives on a schedule tied to time — hours, months, retainers — while the thing you wanted is tied to results. Those two clocks have no relationship to one another. The vendor's revenue is complete the moment the work ships. Yours begins, if it begins, much later.
This creates a quiet, structural problem: the person building your growth engine has no financial stake in whether it compounds. Not because they're dishonest — most aren't — but because nothing in the arrangement makes their outcome depend on yours. Good intentions are asked to do the work that structure should be doing.
And intentions decay under pressure. When a project runs long, when a metric disappoints, when a better-paying client appears — the vendor whose money is already made and the vendor whose money depends on your growth will behave differently. Not because one is virtuous. Because one is exposed.
The flaw isn't in the people. It's in the shape of the deal.
The Perfect Act
There's a word for someone who takes a small position in something larger than himself and moves it toward value everyone shares. In finance, that's an activist — a minority holder who shifts a company he doesn't control, not by force, but by a credible plan and an aligned interest. It works for exactly one reason: when he wins, everyone wins.
Somewhere the word turned ugly. The activist became the raider — uninvited, extractive, in to squeeze and out before the bill came due. The word now carries that baggage almost everywhere it's used.
We are the other kind. The kind the word was always meant to describe, and almost never was. Invited, not imposed. Building, not stripping. Paid over years, not quarters.
An activist is, at root, simply one who acts. The question is only ever what the act is, and whether it's finished in form. Ours is building — and building something that keeps paying after we've stopped touching it, for as long as it works.
That's what "perfect" means here. Not flawless. Finished in form — the thing activism was always trying to become, with the extraction removed and the building left in.
A perfect act, then: to build something meant to compound for someone else, and to profit only out of what it creates.
The Deal
The whole model, stated plainly:
You run a business that's growing. We take a position in that growth. We build the engine to compound it — a real product. You own it. Beyond a capped base that covers our cost, we are paid only out of the growth it creates, for as long as it keeps working.
Three properties do all the work:
We are paid on a delta, never a transfer. Our share comes out of growth that did not exist before we built anything. We are never paid a slice of what you already had. If nothing new appears, nothing beyond the capped base is owed.
You own the asset. Not a license, not a dependency, not a thing that stops working when you stop paying. The product is yours. This is deliberate: an arrangement where we could hold your engine hostage would corrupt everything above it.
The term is long. We are paid for as long as it keeps working, which means we are exposed to whether it lasts. A vendor optimizes for delivery. A partner exposed for years optimizes for durability. That difference is not a matter of character; it's a matter of what the contract does to us.
And one part is fixed, by design. The engine has to be built before it can create anything, and someone pays for that work up front. The capped base does — the builders' time and the tools. It is fixed in your agreement before any code is written, it never grows, and it covers our cost, not our profit. Every software buyer already pays for work before results; a retainer bills it by the hour, with no link to the result and no ceiling. Here it is one capped figure, known in advance, and everything we earn beyond it comes only out of growth that appears.
You don't have to trust an alignment claim. You have to read the deal.
The Front Door
Most firms qualify you. We let you qualify yourself, in the open, before you speak to anyone.
Two doors, mirror images of each other, and both ask the same question in different currencies.
If you own the business: you enter your own numbers — what a product would grow, what share of that growth you'd give, what lift you'd expect. The instrument tells you what the split would look like, and it tells you honestly when the answer is this doesn't fit. Not a lead form dressed as a tool. A verdict you can get without talking to us, including a negative one.
If you build things: the same instrument, reversed. A base you can count on plus a share of what you help build — and it tells you plainly that if you want pay regardless of outcome, this isn't the room.
The symmetry isn't decoration. A business owner who won't share upside and a builder who wants pay divorced from outcome are the same misalignment, seen from two sides. One gate excludes both.
It is not trying to convert you. It's trying to tell you the truth early enough to be useful.
The Scoreboard You Write
Here is the moment everything else depends on, and the one we deliberately refuse to automate.
Before anything is built, we sit down together and write the document that defines what we're building, what number counts as success, and — critically — what that number would have done without us.
That last figure is the baseline. It is the honest counterfactual: the growth you'd have had anyway. Our share is calculated on the gap above it, never on the whole line. Which means the baseline is the single most consequential number in the relationship, and it is set before anyone has an incentive to argue about it, by both of us, in writing.
You write your own scoreboard. We hold the pen with you.
Two things follow. First, nothing is built until that document passes its own test — if the success metric isn't measurable, or the baseline isn't fixed, or the growth being claimed isn't real growth, work does not begin. Second, no one has to be trusted afterward, because the standard was agreed before the outcome was known.
This is slow on purpose. It is the one thing in the entire machine that we would never make instant, because the trust it produces is the slowness.
Who Decides If It Worked
Nobody.
That's not evasion — it's the design. In most arrangements, whether the work succeeded is a judgment call, and the person best positioned to make it is the person being paid. We removed that position entirely.
The standard we are held to is the one written in the document above, fixed in advance, measured against a baseline both parties set. When the period closes, the number either cleared it or it didn't. There is no meeting where we explain why a disappointing result was actually a success. There is no room for that meeting to happen in.
Only structure and data are the judge. Structure decides what the rules are; data decides what happened. Human judgment gets exactly one job — authoring the terms up front, in the open, with you — and is then removed from the verdict.
The test of any alignment claim is simple: who gets to decide whether it worked, and do they get paid either way? If the answer is "we do, and yes" — the claim is decoration.
The Engine Room
Between a signed agreement and a live, measured position there are a great many steps. Qualification, briefing, scheduling, drafting, instrumenting, measuring, reporting, invoicing. Most firms fill that space with people, and the people become the bottleneck and the cost.
We automated nearly all of it — and the reason is not efficiency.
It's that maximal automation concentrates human attention on the one thing that must stay human. Every step the machine can walk is a step where our care isn't needed. What's left is the single judgment nobody should delegate: sitting with you and deciding whether this is genuinely positive-sum, and writing the baseline honestly.
The machine walks each engagement to the door of that decision, and waits there. It cannot open it. Automating that moment would delete the trust the moment exists to create.
So: fast wherever speed is honest — qualification, the call request, measurement, reporting. Deliberately slow at exactly one place. That asymmetry is the product.
The Meter
Once something is live, the measurement isn't a report we write about ourselves. It runs off the product's own data.
The engine we build ships with its own scoreboard attached. What it produced is recorded as it happens, against the baseline you fixed. Nobody assembles a narrative at month's end.
Which produces something we find genuinely elegant: the invoice and the evidence are the same object. The measurement that shows what was created is the calculation of what's owed. There is no separate billing act, because the amount was determined the instant the growth was measured — your agreed share, applied to the gap above your agreed baseline. Arithmetic on two numbers you helped set.
You are never sent a bill you have to evaluate. You watch what's owed accrue, in the open, on the same instrument that shows what was built. Payment isn't argued. It resolves.
One Floor, Seen Whole
Everything above lands on a single view.
One instrument. One clock. One set of rules. What differs between people looking at it is only how much of it they can see — you see your own position; we see the whole book. Same floor, different vantage.
What is not permitted is a second version. There is no internal ledger where the real numbers live and a friendlier one shown to you. The position you see of yourself is, exactly and structurally, the position we see of you. Not as a policy we promise to follow — as a property of how the thing is built.
Three disciplines keep it honest. Every curve is drawn against its baseline, so that what was created is the gap, never the whole line — a chart without its baseline is a flattering picture, and we treat it as a defect. Every figure wears its own freshness, so you know whether you're looking at something live or something last assessed months ago. And the future is drawn as a widening range, never as a confident line, because a confident line about the future is the most common lie in this industry.
The Book
A word about what we are, because it explains our behavior better than any statement of values.
We are not an agency with a novel pricing model. We are closer to a fund that takes operating positions. Each engagement is a position: we put our own work in first; beyond a capped base, we're paid only out of the growth it creates; and we hold it for years. A book of those positions is what the firm is worth.
This has a consequence you should be able to verify from the outside: a position that doesn't work costs us everything we would have earned beyond our cost. A fee-based firm is paid in full for a failed project — the fee cleared. We are paid our capped base, and nothing more. Our survival is a function of how many of these actually work, which is the only kind of alignment that doesn't require anyone's good character.
It also explains why we say no. A book concentrates our capacity in a small number of positions, so each one has to be worth years. Selectivity isn't prestige. It's arithmetic.
What This Does Not Promise
A document like this should be measured by what it refuses to claim.
We do not promise your growth. Nothing here guarantees the number goes up. Markets, timing, and execution beyond a single product all sit outside what anyone can honestly commit to. What this structure guarantees is narrower and more useful: that the question will be measurable, that the scoreboard will be honest, and that beyond a capped base, we are only paid out of growth that actually appears.
We do not promise speed where speed would be a lie. Qualification is instant. Measurement is instant. Building an engine that changes a real revenue line is not, and any firm telling you otherwise is selling the feeling of progress.
We do not promise that this fits you. For a great many good businesses it doesn't, and the front door is built to tell you so quickly and without a sales conversation.
The claim is not that we're better people. It's that we've removed the places where being worse would pay.
The Invitation
Two doors into the same engine.
If you run a business that's growing and you want it to grow faster — and you'd rather cap what the attempt costs and share what gets created — you're not a client. You're the perfect partner. Go put your own numbers in and see whether it fits, including the possibility that it doesn't.
If you build things — and you'd rather own a share of what you make than draw a fee regardless — the same door is open from the other side, and the terms are structurally the same ones. The whole model would be a lie if the people building the engines were paid on a different logic than the engines themselves.
Both gates run the same test, in each side's currency, and both are honest about a negative answer. If the position isn't positive-sum, the instrument says so — and there's no call to request, because there'd be nothing to build together.
Beyond a capped base that covers our cost, we're paid only out of the growth we create, for as long as it keeps working.
That is the whole of it. Everything else in this memorandum is the machinery that makes the sentence true instead of merely nice.