When Building Is Cheap, Focus Is the Whole Job

We've argued elsewhere that volume is free now — that producing working software got commoditized, and judgment became the scarce input. That's the first half, and it's true. Here's the half that bites the founders we work with: when you can build anything cheaply, you build everything. The failure mode flips. It used to be "we couldn't ship it." Now it's "we shipped constantly and nothing decisive happened."

Cheap volume doesn't reward effort. It punishes the absence of focus, faster. A team that ships ten mediocre things a week now feels productive while the one or two things that actually matter sit half-finished underneath the pile.

This is the pattern across the highly technical founders we step in to help: not a shortage of output, a shortage of direction for it. They can do anything, so they're doing everything, pulled in twelve directions at once. The work we bring isn't more building. It's deciding what gets the building.

Most of the value is hiding in a few places

The math here is old. The Pareto principle — roughly 80% of outcomes come from 20% of causes — has held since Vilfredo Pareto noticed 20% of his pea pods produced 80% of the peas in 1896. In a business: a few customers, a few features, a few decisions carry most of the result.

The trap is the lazy reading of it. People hear "80/20" and think it means less effort — do the easy fifth, skip the rest. It doesn't. The 80/20 rule is about causes, not effort: you still pour 100% into the right 20%. The skill isn't working less. It's aiming all of your force at the small set of things that move the number.

Call it value-weighted focus: you find the few initiatives that carry the value, and you spend the lion's share of your capacity — call it 80% — on those, deliberately, before anything else gets oxygen.

Put the boulders in first

The image we keep coming back to is the jar. Stephen Covey's "big rocks" demonstration: fill a jar with sand first and the big rocks never fit. Put the boulders in first, and the pebbles and sand settle into the gaps around them. Same jar, same day, completely different outcome — decided entirely by order.

Most engineering orgs run the jar backwards. The sand goes in first — the urgent-but-small, the tickets, the someone-asked-for-it, the easy wins that feel like progress — and by the time anyone looks up, there's no room left for the boulder that would have changed the quarter. Cheap, fast tooling makes this worse, because it manufactures sand at an incredible rate.

So how do you tell a boulder from a pebble before you've spent the quarter? You score it. We lean on Weighted Shortest Job First — cost of delay divided by job size, an idea drawn from Don Reinertsen's economics of product flow. Rank each candidate by the value it unlocks (and the cost of not doing it) against the effort to ship it. The highest score goes first. It turns a roomful of opinions and a loudest-voice-wins backlog into a sequence you can defend.

What "value" actually means — and the order that matters

Scoring only works if you're honest about what value is. For our clients it has a strict order, and the order is the part most teams get wrong.

First, new revenue — the faucet. The work that opens a market, lands the deal, turns the product into something more people will pay for. Second, protecting revenue and killing churn — the leak in the bucket. This is wildly underweighted: Bain's research found a 5% lift in retention can raise profit by 25% to 95%. Plugging the leak compounds harder than almost anything else you can do.

Cost reduction comes last, and on purpose. If your bucket leaks or your faucet trickles, trimming the cloud bill is rearranging deck chairs — you're optimizing a system that's losing money at the top. We reach for cost only once the revenue and retention boulders are handled. It's the least glamorous truth in the practice and the one that saves the most companies from optimizing themselves into irrelevance.

Here's how that turns into a week of work:

1. Define value as money
   Growth, retention, then cost — in that order. If an initiative can't be tied to one, it isn't a boulder.
2. Score, don't argue
   Run the candidates through cost-of-delay over effort. Let the ranking, not the loudest stakeholder, set the sequence.
3. Load the boulders first
   Commit ~80% of capacity to the top of the list before the pebbles and sand get any.
4. Leave room for sand — last
   Keep-the-lights-on work is real, but it goes in the gaps. It never goes in first.

None of this means you do only the boulders. The lights have to stay on; the small stuff is real. It means the boulders go in the jar first and claim most of the space — and everything else competes for what's left, not the other way around.

The founder from the opening didn't need to build more. He needed someone to stand in the room and say: these two things are the quarter, the other ten are sand, and we're going to act like it. That's the move. When the cost of building drops to nothing, the value of choosing what to build goes up — and it becomes, quietly, the whole job.