Why humans are bad at estimating (and how to get better)
by Bart Etcheverry ✦ 21 October 2025 ✦ 5 Min Read
We all have to estimate. Whether it’s figuring out if you’ve got time to grab snacks before the train, or squeezing in one more task before a meeting. We’re all doing it constantly.
If you work in a product team, you’re probably estimating a little bit more than the average.
And if you’re an agency like us, where every project starts with an estimate, you have to get pretty good at it.
But the truth is, humans are wired to be bad at estimating. We’re optimistic, biased and forgetful (which is fine, as long as it’s not your loved one’s birthday you’re forgetting.)
What really matters is how we build better habits around it.
Here’s how we approach estimation at Planes: three simple ideas to make it more useful, honest, and confidence-building.
1. We’re terrible at predicting the future (and that’s normal)
Even the most experienced teams fall into familiar traps:
Hofstadter’s Law: it always takes longer than you expect... even when you expect it to take longer.
The Planning Fallacy: we underestimate work, even when we’ve done similar things before.
Optimism Bias: we assume things will go smoothly, despite all evidence to the contrary.
That doesn’t mean estimating is pointless. It means we need to treat it as a conversation, not a quick answer.
Estimation helps us talk about how much we know and where we might be wrong.
It’s less about accuracy, and more about confidence: in our understanding of the work, the risks, and the trade-offs we’re making.
The goal with estimation is always confidence. Confidence that we understand enough to make smart decisions.
2. We’re better at comparing than calculating
In a recent internal session, we ran a simple experiment with two jars of jellybeans: one big, one small.
When asked “How many jellybeans are in the large jar?”, guesses ranged wildly.
But when asked “Does the smaller jar have more or less?”, everyone aligned instantly.
That’s relative estimation in action.
Humans are bad at absolute prediction, but great at comparison.
That’s why we use techniques like:
T-shirt sizing (S, M, L, XL) for early scoping
Story points for effort and complexity
Fibonacci sequences to account for increasing uncertainty
These frameworks give teams a shared scale. More importantly, they help us compare, not calculate. Which creates alignment and reduces risk, without pretending we can see the future.
3. We use timeboxing and real data to keep us honest
Estimation on its own doesn’t make us better. Feedback does.
That’s where timeboxing comes in.
By setting clear time boundaries, we give ourselves a defined space to measure progress and collect data.
Rather than aiming for certainty, we create a structure to observe reality:
“Let’s work on this for two weeks and see what that tells us.”
Each timebox becomes an experiment. It helps us see how much we can get done, where we’re blocked, and what slows us down.
We then use real delivery data — velocity, throughput, or cycle time — to calibrate future estimates.
These metrics aren’t about performance or pressure. They’re about creating a feedback loop between what we thought would happen and what actually did.
Timeboxing gives us the rhythm to learn. Data helps us steer with confidence.
Estimating will always involve uncertainty
But by comparing rather than guessing, constraining rather than over-planning, and learning rather than blaming, we can make estimation feel less like a gamble, and more like a shared act of learning.
The goal isn’t to predict the future.
It’s to build the confidence to navigate it together.
