Structural Business Alignment Through Common Interests

Structural business alignment through common interests. Company A needs company B, because of reason Y, and in this way A and B grow, aligning themselves to reach Y. Also the parameters that make A and B collaborate, each and every one of them, which are the most habitual causes and in addition the non-habitual ones, that is to say all of them, and how the companies that appear to solve the needs of B1, B2, B3, B4, B5, B one million, obtain financing simply by being aligned with the interests of that company, making clear that this is a feasible and coherent way of working with respect to the ambition of a common objective. In addition to the cases of different situations but aligned to the same thing.

What follows is that proposition with its figures.

The number that establishes it

In the United States this year, corporate investors account for a record 87.9 per cent of the value of artificial-intelligence venture deals, and artificial intelligence now represents more than nine tenths of all corporate venture deal value.

Read that carefully, because it is not a statement about how much money is being invested. It is a statement about who is doing the investing and why. A corporate investor is not a fund seeking a return. It is an operating company placing money in another company because that company's success makes its own business work better. The cloud providers invest to secure access to models and to commit future consumption of their own infrastructure. The enterprise and infrastructure firms invest to embed capability inside their own products. The largest corporate investor by deal value is the company that sells the hardware all of them depend on.

The practical consequence for anyone building something is the entire subject of this piece. Capital is no longer allocated principally by investors asking whether a business will work. It is allocated by operating companies asking whether a business makes their own roadmap arrive sooner. Being useful to a large company's plan has become a financeable asset in itself, independent of revenue, and it is currently the single most reliable way to raise money that exists.

The parameters: why A collaborates with B

What makes two companies align is not goodwill and it is not a shared vision. It is one or more of a finite list of conditions, and the list is worth setting out in full because most people can only name three of them.

Complementarity. B produces an input to A's product and cannot easily be replaced. This is the ordinary case and the least interesting one.

Opportunity cost. A could build it, and should not, because the engineers who would build it are worth more doing something else. Most outsourcing is not about capability. It is about what the capability displaces.

Speed. A can reach a market eighteen months sooner through B than by itself, and eighteen months is usually the whole of the advantage available.

Demand visibility. This is the most undervalued parameter in commerce. A's roadmap is B's sales forecast. A supplier who knows what a large customer intends to build in two years can invest against that knowledge, and that certainty is worth more than a higher price.

Capacity reservation. A pays in advance to guarantee output that does not yet exist, which finances B's expansion and secures A's supply. The largest example in the world is documented elsewhere in this newspaper: $279 billion of purchase obligations placed years ahead with a manufacturer.

Distribution. B gains access to A's customers without paying to acquire them. For most small companies this is worth more than money.

Standards adoption. B builds to A's interface, which lowers A's integration cost to near zero and makes B the path of least resistance for everyone else inside A's ecosystem.

Risk transfer. A wants an experiment conducted without carrying it on its own balance sheet or its own reputation. B conducts it, and is paid for absorbing the possibility of failure.

Regulatory or geographic access. B holds a licence, a presence or a relationship that A cannot obtain in a reasonable time. This is the whole of the market-entry industry.

Capital access. A invests, and the investment is worth more than the money because it is a signal. Everyone who later looks at B sees that an informed buyer has already committed.

Reputation transfer. B borrows A's credibility. A customer who will not sign with an unknown supplier will sign with an unknown supplier that a known company has endorsed.

Talent and knowledge. The collaboration is a way of acquiring a team or a competence without acquiring a company.

Incentive alignment through equity. A takes a stake in B specifically so that the relationship survives the first serious disagreement. Contracts govern behaviour; ownership governs motive.

The parameters nobody lists

The habitual causes above explain most collaborations. The unusual ones explain the most interesting.

Pre-competitive cooperation. Direct competitors share the cost of something that does not differentiate them: a standard, a safety protocol, a measurement method, a piece of basic research. Each would have paid for it alone and none gains an advantage from owning it.

Shared physical infrastructure. Rivals that compete fiercely for customers share towers, runway slots, payment rails and warehouse networks, because duplicating the asset would make both of them worse off and neither of them better.

Deliberate creation of a second source. A large buyer funds a competitor to its own supplier, on purpose, to ensure it is never dependent on one. The modern processor industry exists in the shape it does because a dominant customer once demanded that its chip supplier license a rival to make the same product.

Cross-licensing. Two companies with overlapping patents agree not to destroy each other, because the alternative is that both spend a decade in court and a third party takes the market while they argue.

Pre-emptive standard setting. An industry organises itself to write a rule before a regulator writes it for them. The cooperation is real, the motive is self-interested, and the result is usually better than either alternative.

Co-opetition. One company supplies a critical component to a direct competitor, because the margin on supplying is certain and the margin on competing is not.

Hostage exchange. Each party deliberately makes itself dependent on the other, so that defection is expensive for both. This is the same instrument used between states and described elsewhere in these pages; it works in commerce for the identical reason.

Different situations, the same alignment

The point that makes this useful is that the companies in an ecosystem do not resemble each other at all.

Around any large platform there are firms doing entirely unrelated things: one writes design software, one manufactures a component, one implements the system inside client organisations, one trains people to use it, one finances the hardware, one insures it, one recruits the staff that operate it. They share no technology, no customers and no business model. What they share is a single dependency: all of them get larger when the platform at the centre gets larger.

That is alignment to the same thing from entirely different situations, and it is why ecosystems grow faster than companies. The platform does not have to fund, manage or even know about most of the firms that are making it more valuable. Each one is solving the need of a different B, and all of the Bs are pointed in the same direction.

It is also why being that kind of company is financeable. The investor is not betting on the firm's ability to create demand. The demand already exists and is growing for reasons the firm does not control and does not have to.

A worked example: a labour marketplace

Perhaps I should cite Trabajo Abierto here, right, and how it can align itself with business interests that want to sell more and pay less in salaries.

And not only that, but to provide a surplus of players waiting for their offers to fulfil them in the short term, without monthly salaries. That, aligned to many companies, equals financing by the ton to address interests in common.

In the terms set out above, this is a B. It does not create the demand it serves and does not need to. That demand already exists, simultaneously, in B1, B2, B3, B4, B5 and B one million: in every company that has work arriving in bursts which cannot justify a permanent salary, and which therefore either goes undone or is done late by somebody already occupied with something else. The marketplace does not persuade any of them that the need exists. It supplies the capacity to meet it on the day it appears.

That is what makes it the same structure as every other example in this article. A B does not invent a market. It positions itself against a demand that somebody else's situation has already created, and grows at the speed of that situation rather than at the speed of its own persuasion.

The alignment being described is between a labour marketplace and the companies that need work done, and the parameters it satisfies are three of the ones listed above: demand visibility, capacity reservation and distribution. What a company obtains is not a discount. It is a different shape of labour.

The figures that make the case are published and are larger than most people assume. The median cost of filling a non-executive position in the United States is reported at $1,300 and of an executive one at $15,000, with average figures from the same body considerably higher at $5,475 and $35,879. The median time to fill a position is thirty-nine calendar days. A recruitment agency charges between twenty and twenty-five per cent of the salary. And the cost of hiring the wrong person is estimated at between thirty and fifty per cent of annual salary, or around $14,900 on average, rising above $250,000 for a senior role once replacement, lost output and disruption are counted.

Those numbers are the friction, and friction is where the surplus in this model comes from. A sixty-day vacancy costs a company tens of thousands in output that was never produced, which both sides are paying for: the firm in lost contribution and the worker in time spent unemployed while the position sits open.

Multiple heads, same matter

Yes, it sounds good. What sells most is having multiple heads on the same matter in a cheaper way and still being beneficial for workers anywhere on the planet.

Compare it with Hetzner and renting ten servers for a short period of time, instead of contracting one at Contabo for one month for the same price. Both are fine, but one solves in one day what the other solves in a month, costing the same, although luckily in this case it is even less costly for the company and more beneficial for the workers, so in the end everybody wins.

That comparison is the clearest statement of the model that exists, and it is exact. The same budget buys either one unit of capacity for a long period or many units for a short one. The money is identical and what is purchased is not: in one case a result in a month, in the other a result on Tuesday. The computing industry has understood this since the cloud existed, and almost nobody has applied it to people.

I do not mean that, but thirty workers on different matters, as in a company.

That thing about ten people not giving birth to a baby in a month refers to a job that has nothing to do with this; it is for something where it especially has to be that person, so it does not apply.

No less important, it aligns with the interests of millions of people all over the planet, spread across countries where the salary is incredibly unfair compared with countries like Switzerland. Switzerland versus Cuba, in salaries for the same work. It is a real discomfort for people all over the planet, who without applying will go, delighted with life, to carry out an open job. Which fits with Hetzner and Contabo: ten servers in three days or one in a month, for the same price.

That disparity is measurable and the figures are published, with one caution that has to be stated before any of them: average wages and legal minimum wages are different measures, and comparing one against the other is the commonest way to exaggerate a gap that needs no exaggeration. What follows keeps them apart.

On average wages, adjusted for what money actually buys locally, Switzerland sits at about $6,096 a month and Nigeria at about $160. That is roughly thirty-eight times, between two averages, on a measure already corrected for the cost of living.

On legal minimum wages, the spread is wider still and the bottom of it is difficult to credit. Switzerland has no federal minimum; the cantons that set one are among the highest in the world, at several thousand dollars a month. The Philippines is around $300. India is around $75 and varies by state. Nigeria is around $51. Cuba's minimum, set at 3,210 pesos by resolution this year, is worth about $4.28 a month at the informal exchange rate of roughly 750 pesos to the dollar. Venezuela's legal base has been frozen at 130 bolívares, about $0.27, since March 2022, with the rest arriving as bonuses that do not count towards pensions, severance or holiday pay, bringing total monthly income to around $240.

Those numbers describe the same hour of human attention priced at four dollars a month in one place and several thousand in another. For physical work tied to a place, that difference has an explanation: the work cannot move. For work that can be transmitted, reviewed, translated, called, catalogued, designed or verified, there is no productivity difference whatsoever. The output is identical. What differs is the postcode of the person producing it, and until recently the postcode was destiny.

The mechanism by which both sides gain is worth stating plainly, because it is what makes the arrangement durable rather than extractive. In a conventional hire, both parties pay for the mismatch: the company pays for hours in which there is nothing to do, and the worker is tied to one employer regardless of spare capacity. Removing the mismatch creates value that did not previously exist, and value that did not previously exist can be divided between both sides. The saving comes out of dead time, not out of wages, which is why the marketplace is not a zero-sum arrangement and why a worker anywhere on the planet can be paid for a task without having to move to the country that is paying.

And there is a discipline inside the model that no owner can escape. A two-sided marketplace is valuable only in proportion to its liquidity, which is the probability that a request is filled quickly. Squeeze the supply side and the available people leave, liquidity falls, and the asset that was being sold to the companies evaporates. Alignment here is not a principle. It is a survival condition.

Where it fits

Not all work suits this arrangement, and a list that pretended otherwise would be useless. Five conditions decide it. The work has to be divisible, so that ten people can do ten pieces of it at once. It has to be verifiable, so that whoever pays can see whether it was done. It has to arrive in bursts, so that a permanent salary cannot be justified. It has to need no embedded context, so that a capable stranger can do it without six months inside the company. And it has to be work a machine cannot simply absorb, which rules out a great deal that would have belonged on this list three years ago.

That last condition is deliberate and it is the one that matters. As set out elsewhere in this newspaper, the price of machine reasoning has fallen roughly a thousandfold in three years, and what it fell on hardest was the transformation of text and images. Translation, transcription, catalogue descriptions, retouching: those were the obvious candidates for a marketplace like this in 2022 and they are the wrong answer in 2026.

What has not moved, and will not, is selling. A machine can write the message and cannot be the person who is trusted. It can find a thousand names and cannot be introduced by a friend. It can describe the product and cannot stand in the room while somebody decides to spend two million. Everything below is that, or directly serves it.

A seller of yachts or aircraft, or an estate agent, can have a hundred open workers instead of one employee. The rest of the list has the same shape.

Where	The selling work	Why a machine does not take it
Property, yachts, aircraft, heavy machinery	Finding and qualifying the few buyers who exist, through a personal network	Already paid on commission, so nothing in the model has to change. The scarce thing is knowing who has the money and being trusted by them, and neither is for sale
Consumer brands entering new cities	Opening points of sale, shop by shop, street by street	Somebody has to walk in, ask for the owner and come back three times
Wholesale and distribution	Getting the product onto the shelf and keeping it there	The second part is the business. A listing that nobody restocks is not a sale, it is a sample
Local services	Street-level and door-to-door selling in a defined area	The customer decides in a conversation that only happens if someone turns up
Market entry abroad	Selling in a country where the company has nobody	A week of a local person replaces a year of guessing, at a thousandth of the cost
Business to business	Obtaining the meeting, in person, with the one person who signs	The gatekeeper exists precisely to stop messages. A person who already knows them walks past it
Trade shows and events	Selling at the stand for the three days it lasts	Demand concentrated into seventy-two hours, after which the need disappears entirely
Dormant customers	Visiting the ones who stopped buying and finding out why	The answer given to a form and the answer given to a face are not the same answer
Demonstration	Showing the product working, in the customer's own place	It is needed precisely because the description did not convince
Franchise and dealer networks	Finding, meeting and vetting the people who will sell it	Choosing who represents your name is the decision nobody delegates to software
Collections and reorders	Going back for the second order, face to face	The first sale is bought with marketing. The second is kept with presence
Retail audit	Checking the product is there, visible and at the right price	Supports every sale above it. Requires standing inside the shop
Lead verification before the visit	Confirming the business exists, at that address, with that owner	A listing can be scraped; whether the door opens can only be established by pushing it
Seasonal peaks	The six weeks in which a year's selling happens	Demand quadruples and then stops, and nobody can hire for that

What does not fit is as instructive. Work that requires accumulated knowledge of one company. Work that cannot be checked without redoing it. Work carrying legal responsibility attached to a specific licensed person. Work that is genuinely continuous, five days a week, every week. And anything that consists of turning one text or one image into another, which no longer needs a person at all.

The media case

You can also talk about other types of business collaboration through kindred ambitions, in the area of quality media representation, in which certain companies use articles in advertising to widen the perspectives that customers obtain about them, even doing it through traditional media such as the giants of paid advertising. That is to say, articles flowing paid traffic on those networks weigh differently from a company talking about itself. However, the interests of the medium must also be seen to be aligned with the interests of the company and to fit with what they are; otherwise they would end up destroying each other mutually. So they are complex matters to decide, but it is very clear when it fits perfectly and really also very clear when it does not. The issue is those that are on the balance right in the middle. However, after an exhaustive analysis before starting, it is simple to see who fits the parameters and who does not, in order to play in the appropriate way from beginning to end, always.

But I said that the companies take the article and advertise it themselves with access to the article, and in the article goes the URL. That is why I said paid advertising on traditional channels adapted 100 per cent to their traditional cash flow.

There is a mechanical reason why this can outperform sending the same money to the company's own page, and it is the auction described in these pages. The platforms do not simply sell to the highest bid; they weight the bid by the probability that the audience responds, because an advertisement people engage with is worth more to the platform than one they ignore. An article holds attention longer than a product page. Longer attention improves the predicted response, and an improved predicted response lowers the cost per result. The advertiser may therefore pay less for the same outcome, which turns the argument from a matter of taste into a matter of arithmetic that can be settled in a week with a small budget and two destinations.

The condition stated above is the one that decides everything, and it is not sentimental. A medium's only asset is the reader's belief that it is not for sale, and that belief is destroyed by precisely the thing that converts it into cash fastest. A publication that sells its credibility collects it once. One that protects it rents it indefinitely. That is why the fit has to be decided before anything is agreed, and why what is paid for is labelled as paid, which in most jurisdictions is not optional and in all of them is cheaper than the alternative.

Winning twice

Agreeing also that one possibility can be key for another. For example, a company receives an article and advertises it, and in that article it obtains autonomous cash flow for them directly, and at the same time obtains visits for a project that aligns with the interests of that same company, generating a double gain. In cases, for example, a person who advertises an article in a medium that also advertises another medium that obtains labour through sales, which the initial company can take advantage of to sell its own products, winning twice at the same time, aligning itself to the objectives of one and the other, building reputation, cash flow from direct ads, and cash flow from wholesale labour only at the price of success, among other cases that can be equally favourable, due to the main theme of the article: the alignment of interests between companies A and B.

Besides, Trabajo Abierto being promoted just by living in the same house is also beneficial.

That last arrangement has a name in the trade and is as old as newspapers: house advertising. Every publication in the world has always filled the space it did not sell with promotion of its own products, and the reason is the economic argument for a media group existing at all, which is that traffic is bought once and used twice. The advertiser pays to bring a reader to an article. The same visit, at no additional cost, exposes whatever else the group operates. One invoice, two results.

The structure described above is three gains from one act: reputation, which is slow and compounds; direct advertising revenue, which is immediate; and labour placed at a price paid only on success, which is the component that removes the buyer's objection entirely, because nobody has to be convinced of anything in advance.

What this establishes

That alignment is a feasible and coherent way of working with respect to the ambition of a common objective, and that it is currently how most capital is allocated in the fastest-growing sector of the economy.

The companies that appear to solve the needs of B1, B2, B3 and B one million do not need to persuade anybody that demand exists. It exists already, it is visible in somebody else's published roadmap, and the investor can read it. That is the whole of the advantage, and it is available to anyone willing to look at what a large organisation has said it intends to do and to build the thing it will need in order to do it.

Where this ends up

How Trabajo Abierto could change the life of millions of people in the early future, who live in countries where their salaries give them a deplorable life, economically speaking, and likewise help every company on the planet that has an open job to offer.

From this premise two points can emerge, and both arrive at a single point: economic growth.

One. Ten Hetzner servers in three days instead of one a month, and clearly I am not talking about computing servers. They could sell many times more than what they sold before, and paying only for what is sold: fewer fixed costs, more cash flow for every company on the planet that fits, equals economic growth.

Two. The open workers can obtain capital money to create new businesses now that they know how to sell, being probably millions of people who receive an abysmal difference between the day's pay of Trabajo Abierto and their minimum wage in their country. They could create multiple companies, startups and even unicorns that drive the planetary economy, above all if they read the article From My Brain to You: The Capitalist System and How to Tame It, or they are simply clever from the factory.

Until next time.

Sources

Capital. The figure of 87.9 per cent of United States artificial-intelligence venture deal value attributable to corporate investors in 2026, and the observation that artificial intelligence exceeds nine tenths of all corporate venture deal value, are from PitchBook's quarterly analysis of corporate venture activity. The characterisation of corporate investor categories, including cloud providers investing to secure model access and future infrastructure consumption, is from the same source.

Hiring costs. Median cost per hire of $1,300 for non-executive and $15,000 for executive roles, and the median time to fill of thirty-nine calendar days, are from the Society for Human Resource Management's 2026 benchmarking report; the average figures of $5,475 and $35,879 are from the same body's 2025 study and differ because one uses medians and the other averages. Agency fees of twenty to twenty-five per cent of salary, the estimate of $14,900 for an average mis-hire and the range of thirty to fifty per cent of annual salary are industry benchmarks reported in recruitment research.

Wages. Average monthly wages adjusted for purchasing power are from compilations of International Labour Organization and national statistical data. Minimum wage figures are the legal rates in force in 2026; Cuba's is set by Resolution 14/2026 at 3,210 pesos and converted at the informal exchange rate, and Venezuela's legal base has been frozen at 130 bolívares since March 2022, with additional income delivered as bonuses outside the wage base. Switzerland has no federal minimum wage; the figures cited refer to cantonal minimums.

Advertising mechanics. The weighting of auction bids by predicted engagement is the published operating principle of the major advertising platforms. The figures for global digital advertising totals and platform concentration referenced here are set out, with their sources, in this newspaper's earlier reporting.