We build the same things we put money behind.
Three lanes: Build, Buy, Back. All three are held for the long term, and both geographies meet the same standard. What follows is how each lane works in practice, not what we promise.
We build the things that move the numbers.
With you.
Revenue and margin are the point; AI is usually the tool. We work hands-on with a small number of companies to find where the money is being lost, then build what closes the gap: software that runs every day in the real business, the changes to who does which job, the weekly and monthly routines that keep it working, and a written plan the team can follow after we leave. If nobody’s job changes around the system, margin does not move. We build both halves and hand back something the team can run. In Kenya the same work starts one layer lower: the problem is usually the records the business runs on, not the software sitting on top of them.
Our own.
We also build and own products of our own: new businesses where we see a gap worth filling. The same discipline applies, with our own money and our own risk. This is the clearest proof we can offer that we build what we recommend.
We build it to keep running after we leave.
| Layer | What gets built | How you would know it worked |
|---|---|---|
| Economics | A one-page map of where the money is being lost and what a fix is worth. The size of the work we take on is set against that number. | Every decision after that traces back to the number on the map. |
| Systems that run every day | Software that carries out the work on its own, plus what it needs around it: a dependable way to find the right information, clean data, tests that check its answers, monitoring while it runs, and one named owner. Enough controls and records that a senior executive will sign their name to it. | It is still running twelve months later, without us in the room. |
| People and routines | Who does what, who decides what, and the standing meetings around the system. Reworked, written down, and practised with the team before we leave. | The team runs the system on Monday; the meeting where it used to be argued about no longer exists. |
| Finance line | The route from the system to a line on the P&L, with the person who owns those accounts in the room. | Finance signs off on the number and adds no footnotes to it. |
When we buy, we take the wheel.
We buy small and mid-sized operating businesses, whole or in part, with our own money. We look for companies that last and produce cash, where hands-on operating work and better systems add value year after year.
The deal itself does not interest us. It matters only if the years of running the company afterwards are worth having.
We own few companies on purpose. We would rather own three companies we understand completely than fifteen we only keep an eye on. There is no fund behind us, so there is no pot of money we have to spend by a set date and no pressure to look busy in any given year.
In Kenya the filters are identical and we spend longer checking before we commit. We do not buy from an owner we met because the business was for sale.
We buy to own, not to resell. The length of the hold is the point.
We back operators we would trust to build the business themselves.
We put our own money behind founders and operators building businesses that use AI to do real work. Usually people we already know, in businesses we understand well enough to be useful in a bad quarter. Someone else is driving. We sit in the passenger seat with money and the experience of having built one before, including the parts that went wrong.
We are not a fund. We do not run a program, we do not do deals in batches, and we do not write checks just to put money to work. What we bring is small and patient, and we mean to be the call an operator actually takes when something is not working.
In Kenya that rule does the most work of any filter we have: operators we knew before the raise, not operators we met during it.
Our money can wait, and nothing forces us to sell.
The second geography, described plainly.
Sub-Saharan Africa is not a separate strategy and not a separate standard. It is the same three lanes, run where we already hold relationships that predate any deal, with a shorter list because distance is real. We count the cost of that distance instead of talking around it.
What we bring that a United States holding company would otherwise have to hire: relationships older than the transaction, and a working read on which local businesses are real. What we bring that a local buyer often does not: books closed every month, a weekly operating review, and reporting a lender would accept.
What it looks like.
- Whole companies, or large stakes in companies that already produce cash, with an owner we can speak to directly.
- Build work where mobile phones already changed how business gets done and the systems inside the company never caught up.
- Operators building there who we can be useful to in a bad quarter, not only in a good one.
What it is not.
- Not development finance, donor money, or a grant program.
- Not an impact fund. The numbers have to work on their own.
- Not a survey of the continent. One market now, and we add another only when we can name the three people we would call there before any money moves.
The standard is the same; the list is shorter.
| United States | Home base, and most of the operating work. Build work with companies, businesses we have bought, and operators we have put money behind. |
|---|---|
| Kenya | Being built now, not owned yet. A shorter list, more time spent checking before we commit, and people we knew before the business was for sale. |
| What holds | A business we can understand, real customers, cash we can trace, and years of work after the deal that are worth staying for. |
| What changes | The pace. Distance, currency swings, and how a company is governed are counted as real costs before we commit, not assumed away. |
| Adding a market | Only when we can name the three people we would call in it before committing capital. Note 04 sets out the test. |
We hold what we build.
Holding for the long term is not a fourth lane. It is the one thing all three have in common, and the reason we can do all three at once without them pulling against each other.
A fund has a deadline. The deadline is not a detail: it decides which repairs are worth making. Anything that pays off only after the fund has to sell becomes, sensibly, the next owner’s problem. Most of the work that makes a company genuinely durable sits exactly there.
We have no fund and no date by which we must spend or sell. How long we hold is our choice, not a rule set for us.
Four conversations, in this order.
This is not a formal process. Most conversations end at the third one, which is a good outcome for both sides.
The economics.
What the business does, who pays for it, where the money is being lost, and what a fix is worth. Nothing needs to be prepared in advance. If a deck is needed to explain the business, that tells us something too.
The constraint.
The one thing that, if fixed, changes where the business ends up. Something in how the company runs day to day, named specifically, rather than a worry about survival or a complaint about culture.
The shape.
Whether this is build work, buying part or all of the company, or putting money behind you while you run it. Most conversations end here.
The terms.
Short, plain, and written down. We have never needed a long agreement to describe honest work, and a long one has never fixed a bad fit.