← All notesSTRATEGY·17 Jun 2026·4 min

Two king-size beds and other planning mistakes

A renovation story thats really about software. Good planning isnt predicting how it goes — its deciding what youre building before you build it.

I was flat-hunting again, and not by choice. The place I'd been renting got sold out from under me — a long story with a short ending: I had to move. So one afternoon I went to view a new one, and my first impression was genuinely good. New building, never lived in, two proper rooms, handsome furniture, the lot. But. There's always a but, and this one was wearing two king-size beds.

I should explain. Renting somewhere decent in Ukraine, long-term, is a quiet art. You either pay handsomely or you end up in the part of the city the maps gave up on — fine if you drive, except back then I didn't, and in a city as walkable as Lviv, taking the car to work instead of your own two feet felt like a small crime against the place. So the search was narrow to begin with. And most of what was on offer looked like a website from 1998: colours that made your eyes water, banners that made you grieve. You started to wonder whether it was you or them.

Pencil-style rendering of a single bedroom holding two king-size beds, a small framed mountain print, and a comically small fridge by the door.
Fig. 01 — Two king-size beds, one bedroom, no idea who it was for. A floor plan that couldn't decide what it wanted to be.

The apartment that couldn't decide what it was

This last flat had a 24-square-metre bedroom with not one but two king-size beds. Holy cow, you're thinking — and you'd be right. The living-room walls were painted a committed grey, and for ninety-six square metres, the fridge was a shy little thing that clearly expected fewer guests. None of it added up, until the owner explained.

It had been built to let short-term — to people on work trips, to tourists. And for that, the logic held. More beds, more heads, more bookings. Tourists have famously adventurous taste; they'll happily accept grey walls and a mini fridge the way they'll accept fried ice cream or sheep's brains. Then, near the end of the renovation, the owners decided the place was too nice to rent by the night, and switched to long-term. A pivot, made late, with the walls already up. What was left was a lovely, almost-cosy flat for one person or a couple — furnished, immovably, for six.

Top-down floor plan of the bedroom with two king-size beds wall-to-wall, a labelled tiny fridge, and 'guest? office? gym?' question marks where the leftover space should be.
Fig. 02 — Two king-size beds wall-to-wall, a fridge sized for guests who were never coming, and the only honest dimension on the plan: barely any floor left to stand on. Guest? Office? Gym? The plan never said.

A pivot is cheap on a whiteboard and expensive once the walls are painted. The later you turn, the more of the build you turn against.

The same floor plan marked up in red with MOVE and DEMO annotations, showing the cost of changing the layout after the walls were built.
Fig. 03 — The same plan, marked up after the walls were up: MOVE this, DEMO that. Every red line is a decision that should have been free on paper and now arrives with an invoice.

Why the beds mattered

You could bin one bed, of course. But then you were left with the harder problem: the hole it left. Too much floor for a bedroom, so maybe a workspace — except a desk where you sleep means you'll end up sleeping at the office, so now it needs partitioning, which means a redesign, which means clearing the room, doing the work, and paying for it in both time and money. And renovation, the world over, runs to the same script: longer than promised, dearer than quoted, and a genuine pain in the backside. One late decision, and every later one inherited it.

01

Deliberate gets cheaper over time

Decide what the place is for before the walls go up, and every later choice falls into line. The plan does the expensive thinking once.

02

Accidental gets more expensive

Build first, decide later, and you don't get to undo the wall. You get to demolish it, then rebuild — at full price, with the clock running.

It's exactly the same in software

Because of course it is. Redoing, redesigning, unpicking a decision that turned out wrong — it's the same demolition job, the same overrun, the same bill arriving twice. This is why the boring part up front earns its keep: the thinking, the clear goal, the honest answer to what is this actually for, settled before the first line of code is written. Get that right and the rest of the build leans with you. Get it wrong and you're the proud owner of a one-person flat with two king-size beds — charming, well-made, and fighting itself in every direction.

You can do that thinking yourself — it's mostly a matter of asking the awkward questions before they get expensive. Or you can talk to someone who's measured the room before. That, as it happens, is the bit I do.

Still with me?

Weighing a rebuild, a rescue, or a roadmap?