Welcome to the August subscriber edition. As I always say, thank you for being here, and for your continued support of this newsletter. I try my best to bring you something novel and, perhaps, a little different to most technology-focussed newsletters, often dipping into other fields to do so.
With that in mind, this month’s article comes from an unusual source: a set of investment letters written by a fund manager most (technology) people have never heard of. Nick Sleep ran the Nomad Investment Partnership for thirteen years, returned 921% before performance fees by holding just a tiny handful of stocks, and then shut the fund down and walked away.
His letters contain an idea that I think fundamentally changes how we should think about focus in all areas of our life: not just that it matters, but why it matters via compounding, and what shape that compounding takes. I first learned about Sleep’s letters from episode #365 of the Founders podcast, and since read the complete Nomad letters, which are actually being published in book form later this year by Stripe Press.
Back to this newsletter. I’ve written before about constraints, bottlenecks, and the single prioritised list as tools for ensuring focus.
This article comes at this subject from a different, but empowering, angle: the returns from sustained focus aren’t linear; they accelerate. The shape of focus is a convex curve, and most engineering organisations never reach the steep part because they keep resetting it. I hope you see that many of my recent articles are preaching the same message!
Here’s what we’ll cover:
The investor who did almost nothing. Nick Sleep, the Nomad Investment Partnership, and the idea of “destination analysis.”
The shape of the focus curve. Why sustained attention produces accelerating returns, not just better ones.
The cost of resetting the curve. Why context switching keeps you trapped in the flat part of the curve, and what it really costs.
Destination analysis for teams. Applying Sleep’s framework to how engineering teams invest their attention, and why it matters.
The manager’s dilemma. Why managers are structurally forced to context-switch, and what to do about it.
Protecting the curve. Practical guidance for staying focused long enough to reach the steep part of the curve.
If you’d like to dig deeper into related themes, here are some articles from the archive:
The beauty of constraints argues for adding constraints to unlock unconventional thinking. This article explains why the focus that constraints enable produces accelerating returns.
One bottleneck at a time applies the Theory of Constraints to sequential focus on the binding constraint, a cousin of the convexity argument.
One list to rule them all shows how a single prioritised list acts as a structural mechanism for protecting focus.
Slow down to speed up explores why patience and deliberation pay off, which is another way of staying on the curve rather than resetting it.
Scope? How about thoroughness? flips the usual scope conversation on its head, which is the convexity argument applied to how teams think about delivery.
If you’ve been reading the free edition this month, you’ll have seen the focus theme building, so let’s go deeper together.
The investor who did almost nothing
In 2001, Nick Sleep and his partner Qais Zakaria launched the Nomad Investment Partnership. Over the next thirteen years, the fund returned 921%(!), compared to 117% for the global market index.
They achieved this by doing something that most professional investors find almost physically impossible: they bought a small number of companies and then did almost nothing. No buying, no selling, just holding.
By the end, Nomad’s portfolio was concentrated in a handful of businesses, dominated by Amazon, Costco, and Berkshire Hathaway. For comparison, the average mutual fund manager held a stock for around eleven months, a figure Sleep took from Vanguard founder Jack Bogle. Sleep and Zakaria simply held indefinitely. They studied their companies deeply, understood where those businesses were heading, and then sat with their convictions for many years at a time.
It worked.
Sleep called this approach “destination analysis.” Rather than fixating on what a business looked like quarter by quarter, he asked a different question: where is this business going to be (i.e. the destination) in ten years if the things I believe about it are true?
As he wrote in the Nomad letters, “destination analysis is consciously central to how we analyse businesses these days. It helps us ask better questions and get to a firm’s DNA.”
This wasn’t necessarily easy. The psychological challenge of holding firm and not following the crowd was immense (and we’ve seen how erratic the stock market can be recently), and Sleep acknowledged this. His remedy was counterintuitive, and as he put it, in a typically British way: instead of continual frantic activity, “there is a lot to be said for gentle contemplation.”
So why does this matter to anyone building software? Because the dynamics Sleep identified in investing map almost perfectly to how engineering teams allocate their attention.
Most organisations treat their team’s focus the way the average mutual fund treats its portfolio: constant rebalancing, quarterly reshuffles, and new priorities popping up every planning cycle. If you go against the crowd and focus, therefore doing the opposite, I’d argue the same compounding logic applies, exactly as it does in Sleep’s portfolio.
The shape of focus
The insight buried in Sleep’s letters isn’t just that patience pays off. It’s about the shape of the returns. When you hold a great investment for a long time, the returns curve ever more upward, accelerating as compounding takes hold rather than growing in a straight line.
Focus works the same way. When a team starts working on a new problem, the early period is flat. You’re loading context, understanding the domain, building mental models, and making initial progress that feels slow relative to the effort invested. That flat stretch is the frustrating part, and typically we can lose our nerve and switch to something else too soon.
But if you stay with it, something changes. The team starts to see connections that weren’t visible in the initial period of a project, and domain expertise compounds: each new insight builds on previous insights, and the rate of progress accelerates. The same shape shows up elsewhere, like the iceberg of uncertainty where most of the work happens invisibly beneath the waterline, or the long, long, painful slog before a breakthrough suddenly appears.
Hour sixty of focused work on a problem is worth more than hour one, not because you’re working harder or better, but because you’re working with a richer mental model. Week twelve is worth more than week one for the same reason. If you’ve ever been an engineer deep in a problem and had someone tap you on the shoulder, you’ll know the feeling of everything falling out of your head at once. The same applies over longer timeframes.
This is the convex curve. Early effort produces flat returns, and sustained effort produces accelerating returns. The steep part of the curve, where the real value lives, is only accessible to teams that stay focused long enough to reach it, and this is where we often fail.
You might reasonably point out that “convex” is a metaphor, rather than an accurate measurement, and yes, of course; I understand. However the evidence from cognitive science (e.g. Sophie Leroy’s research on attention residue, which we’ll come to shortly) and from the investing world tells the same story: sustained attention compounds in ways that fragmented attention simply can’t.
Consider a concrete example. Imagine two teams, each with the same number of engineers and the same talent. Team A spends six months focused on a search problem: they study the domain, iterate on their approach, and develop deep expertise in ranking algorithms, query understanding, and their product’s data patterns.
By month four, they’re making progress that would have been impossible in month one. Team B spreads the same six months across six different initiatives, spending one month on each: loading new context, making initial progress, and moving on before the compounding begins.
Team B ships a higher quantity of work, but you can be sure that Team A’s single feature will be really good.
Cal Newport’s Deep Work popularised the distinction between deep and shallow work, and it’s a useful starting point. But the convexity framing adds something Newport doesn’t quite capture. The returns from sustained attention accelerate, which is a stronger claim than simply saying deep work is “better” than shallow work.
Newport asks whether you’re doing deep work or shallow work. A better question could be how long have you been on the curve?
