Why almost every plan costs more than expected - the dreaded budget overrun.

What a Sydney swimming pool can tell us about the assumptions underneath a business case

In August, the North Sydney Olympic Pool reopened after five years of being closed. It's a stunning building in a spectacular spot, right under the Harbour Bridge, and by all accounts the locals are pretty happy to have it back.

It's a well documented story but definitely one worth reflecting on (that’s not a pool pun)! But it's also genuinely useful one for anyone thinking about business planning assumptions in Australia right now.

The redevelopment was originally approved at a budget of around $28 million. By the time work actually started in March 2021 that had become $58 million, with the pool due to reopen in late 2022 after what was meant to be a two-year closure. The final figure landed at roughly $122 million, and the doors opened on 7 August this year after being shut for 1,986 days.

It would be easy to read this as a story about people getting everything wrong and certainly some things were clearly missed, like the roof design that had to be dismantled and rebuilt. 

But, a lot of what went wrong was also the sort of thing you only really find once you've started - a heritage building from 1936 that turned out to be in worse shape than anyone thought and only became clear once you start peeling off the layers. Then there’s the disputes (over the previously mentioned steel roof) that took years to work through and that you couldn’t have predicted.

I mean, nobody sets out to be off by that much! But, what we can actually see from this is something that we see regularly and within projects a lot smaller than this one. A significant budget overrun is not uncommon.

The most reliable bias in business planning

Optimism bias is the well-documented (and human) tendency to plan for the version of events where things go more or less to plan. It doesn’t take into account the chance of a disaster or a mess. It’s just the plan.

It isn't down to inexperience or incompetence. Even the most carefully planned projects anywhere in the world run well over, because the bias sits underneath the planning rather than in it. You can be diligent, careful and completely honest with yourself and still end up with a forecast that assumes nothing significant goes wrong. Which, at the point you're building it, nothing has.

The pattern is pretty consistent. It’s hard to make accurate estimates when you're making them at the moment you know the least. Contingency gets set as a percentage of what's already been costed, which means it can only ever cover the things you've already thought of. And the longer the timeline, the more room there is for something to shift. It’s especially difficult in this current political climate when things are changing almost daily.

We see the same shape in businesses of every size and across different sectors. When a new site is built or a new system is implemented. Or perhaps a new product line that looks straightforward right up until it doesn't.

Where we see it happen

In our experience trouble can appear in three places.

First is the timeline. Nearly every plan assumes a fairly smooth ramp-up. In practice there's a stretch where the new thing costs money and isn't yet contributing much, and that stretch is almost always longer than modelled. A new site takes time to find its customers or new staff to get up to speed. Or similarly, a new system takes a while before anyone's genuinely faster using it than they were before.

Second is the base case. Most business cases model one scenario, and it's usually the good one. There may not be a version of the plan that answers the more uncomfortable question - what if this takes eighteen months instead of nine, and costs 30% more? Does the business still work?

Finally, the knock-on. This is the one that gets missed most often. When something runs over, it doesn't just cost more money. It eats attention, holds up other decisions, and uses capacity that had been earmarked for something else. In North Sydney's case, the overrun contributed to the council carrying some $56 million in debt and deferring other work it had planned - like work on schools and community centres. That's really the point: the real cost of a budget overrun is not just the overrun itself.

So what actually helps?

None of this is an argument for pessimism, or for not doing things. It's just an argument for being honest about what you don't know yet.

The businesses we see handle this well tend to do a few things quite consistently.

They always do some in-depth research. What similar projects have actually cost them before, rather than starting from what this one should cost. Their own history is usually the best guide available and or perhaps there’s a business nearby that’s done something similar that you can get data on.

They model more than one version. Not an elaborate set of scenarios - but a realistic downside, and an honest answer to whether the business still holds up in it and where it can’t.

They stage the decision rather than making the whole thing at once, so there are natural points to stop, have a look and adjust, instead of one big commitment made at the moment of least information.

And they go back to the assumptions as things move along. A forecast built in January and never touched again isn't really a forecast by June. It's a record of what you used to think and doesn’t take into account things that have changed, material prices that have increased etc.

Why it matters a bit more in Australia right now

Australian businesses are making these calls in a genuinely demanding environment. Occupancy costs are high, labour is expensive and in some sectors hard to get hold of at all, and borrowing costs more than it did when a lot of current plans were first drawn up.

That combination narrows the margin for error. A project running 20% over is uncomfortable in a good year. In a tighter one it can dramatically reshape what the business is able to do for the next two.

And it's probably worth saying this straight up too… None of this means the decision itself was wrong. The pool is open, it'll be there for decades, and the original problem was real and needed solving. Plenty of projects that run over are still absolutely the right call.

The question was never really whether to do it. It's whether you had a clear idea what you were signing up for at the point you said yes.

Got something similar in the pipeline, or already halfway through one? We'd love to hear about it — get in touch, we're always up for a chat.