In February, a small marketing agency decided to refresh its own website. Everyone agreed it was a two-week job, three at most. The copy was mostly written already, the design just needed tidying, and the developer was a friend.
It’s now May. The website is nearly finished, as it has been for about six weeks.
Nobody did anything particularly wrong. Client work took priority, as it should. The friend went on holiday. The photos needed reshooting. A page nobody had thought about turned out to be essential. Each delay was reasonable on its own. Together, they turned a fortnight into a season.
If this sounds familiar, the comforting news is that it’s not a personal failing. It’s one of the best-documented patterns in psychology, and it affects students, governments and construction firms just as reliably as small businesses. The more useful news is that there are simple ways to plan around it.
One Like insight: We plan for the project we imagine. We deliver the project that actually happens.
The planning fallacy
The psychologists Daniel Kahneman and Amos Tversky first described the planning fallacy in 1979: our tendency to underestimate how long tasks will take, even when we know similar tasks have run late before. In 1994, Roger Buehler, Dale Griffin and Michael Ross tested it in a series of studies published in the Journal of Personality and Social Psychology.
The best known involved 37 psychology students writing their final-year theses. On average, they predicted they’d finish in 33.9 days. Asked for a worst-case estimate, “if everything went as poorly as it possibly could”, they said 48.6 days. In reality, the average thesis took 55.5 days. Only about 30% finished within their own prediction, and on average the students took longer even than their worst-case estimates.
The researchers found that when people plan, they mostly imagine the steps of the current task going smoothly. They rarely think about how similar tasks went in the past, and when they do, they tend to explain away earlier delays as one-off bad luck. Intriguingly, other people’s estimates of how long we’ll take tend to be less optimistic than our own. Outsiders don’t share our hopeful story about how this time will be different.
Thirty-seven students aren’t a business, and a thesis isn’t a product launch. But the same pattern has been found in group projects, tax returns and large infrastructure projects. Anyone who has renovated a kitchen will recognise it.
Making better estimates
Researchers have tested several ways to counteract the planning fallacy. Three are especially practical for small businesses.
Look back before you look forward. Instead of imagining how this project will go, ask how long similar projects actually took last time. Kahneman and his colleagues later called this taking the “outside view”, and in large-scale planning it’s developed into a formal method called reference-class forecasting. In a small business, it can be as simple as checking your diary or invoices to see how long the last website, fit-out or product launch really took.
Break it into pieces. Studies on what researchers call “unpacking” have found that listing every step of a task, including the dull ones such as approvals, waiting for suppliers and fixing mistakes, tends to produce longer and more realistic estimates. Many delays hide in steps nobody bothered to write down.
Ask someone else. Because outside observers are less optimistic about our timelines than we are, a colleague’s estimate is a useful reality check. If they say six weeks and you say two, the truth is probably closer to them.
[Infographic placement] Planned versus actual. A simple horizontal bar chart from the 1994 thesis study. Best-case estimate: 27.4 days. Expected estimate: 33.9 days. Worst-case estimate: 48.6 days. Actual average: 55.5 days, highlighted. Beside it, three small icons with the fixes: Look back, Break it down, Ask someone else. Caption: Buehler, Griffin and Ross (1994): on average, the actual time exceeded even the worst-case estimate.
Planning for the project that actually happens
None of this means abandoning ambitious deadlines. It means building them on evidence rather than hope. Before you commit to a date, especially one you’ll share with a client, take five minutes to look at how long similar work took last time, list every step including the waiting, and ask a colleague for their honest guess. Then add a buffer for the things nobody can predict, because there will be some.
It also helps to separate the date you aim for internally from the date you promise externally. Aim for the optimistic version if it motivates the team. Promise the realistic one. Customers rarely complain when something arrives early.
A few questions can help:
- How long did the last similar project actually take, not how long did we think it took?
- Have we listed the waiting, approvals and fixes, or only the main work?
- What would someone outside the project estimate?
- Is the date we’re promising based on evidence, or on what we’d like to be true?
Which brings us back to the agency and its two-week website, now entering its fourth month. If they’d checked how long their last client website took, about ten weeks as it happens, they’d have planned for something closer to that. The work wouldn’t have been any faster. But nobody would have spent three months feeling behind on a project that was, in truth, roughly on time.
Sources
- Roger Buehler, Dale Griffin and Michael Ross, “Exploring the ‘planning fallacy’: Why people underestimate their task completion times”, Journal of Personality and Social Psychology 67(3), 1994 (DOI 10.1037/0022-3514.67.3.366). Figures checked against Donders Institute’s summary and other secondary sources.
- Daniel Kahneman and Amos Tversky, “Intuitive prediction: Biases and corrective procedures”, 1979 (origin of the term).
- Justin Kruger and Matt Evans, “If you don’t want to be late, enumerate: Unpacking reduces the planning fallacy”, Journal of Experimental Social Psychology 40(5), 2004 (general reference for the unpacking finding).
