Picture an ostrich with its head buried in the sand, convinced that if it can’t see the danger, the danger isn’t real. Now picture a founder doing the exact same thing with a project that isn’t going to plan. That’s the idea at the center of this episode of Brains Byte Back, where Dr. Atif Ansar breaks down why smart, capable people avoid the bad news they most need to hear.
Early in his career, Dr. Atif set out to prove that big dams were a smart bet for developing countries, fully expecting the numbers to back him up, and instead the data told him the opposite. Budgets doubled, timelines stretched out for years, and by his own admission he almost didn’t want to see it.
“If I had not approached it with the kind of emotional neutrality that, as an academic, you are almost drilled to approach it, I would have almost wanted to miss that data,” shared the Oxford researcher. If someone trained to stay objective can feel that pull, the rest of us don’t stand much of a chance.
From there, Atif and his colleagues at the prestigious university built a dataset of 16,000 megaprojects, spanning everything from dams and railways to Olympics and space missions, and the pattern was as consistent as it was brutal. Projects run late, blow through their budgets, and deliver less than what was promised, and almost none of them hit all three targets at once.
So why do intelligent teams keep walking into the same wall? This is the ostrich effect, the very human habit of avoiding information you already suspect is bad. Atif summed it up when he said the instinct is to believe that “if I don’t see it, the risk doesn’t exist any longer,” which is really just the billion-dollar version of the credit card statement you refuse to open after an expensive trip.
The uncomfortable twist is that founders are the worst offenders, because as Atif explained, entrepreneurs “heavily suffer from optimism bias, potentially more than anybody else.” The same optimism that lets you start something hard is what makes you flinch from checking how it’s really going.
In this conversation, you’ll learn how to catch yourself doing it, how to find your real baseline without guilt or blame, the way a lifter has to accept the weight actually on the bar before loading on more, and why the fix has nothing to do with a fancier dashboard.
He also provides a practical tip: founders should start questioning greens and start supporting reds. The metrics glowing green month after month are the ones that deserve your suspicion, while the red ones everyone would rather avoid are exactly where your attention belongs.
🎧 Listen now, and check out Dr. Ansar’s work at foresight.works.
Find out more about Dr. Atif Ansar.
Learn more about https://www.foresight.works/
Reach out to today’s host Erick Espinosa – [email protected]
Get the latest on tech news – https://sociable.co/
Leave an iTunes review – https://rb.gy/ampk26
Follow us on your favourite podcast platform – https://link.chtbl.com/rN3x4ecY
Transcript:
Dr. Atif Ansar: My name is Atif Ansar. I wear two hats. I’m co-founder and executive chairman at Foresight Works. We are an AI startup for speeding up the construction of big projects. I’m also an academic at Oxford University where I’ve worked for over 15 years looking at the success and failure of big projects, and that informed the startup that I’ve been working on for the past eight years.
Erick Espinosa: Amazing. Thank you, Dr. Ansar, for joining me on this episode of Brains Byte Back, especially on the other end of the pond. I don’t know how the weather is there in England right now because I know the last little while you guys were suffering in the summer with some brutal heat. How has it been so far?
Dr. Atif Ansar: It’s been great. I mean, it’s been a beautiful summer in many ways, although brutal heat is indeed the case.
Erick Espinosa: Yeah, I imagine. So I kind of want to start this in terms of maybe like a personal thread because you do mention that you run the Oxford program on sustainable mega projects, and basically kind of you study why big billion dollar projects kind of fail. And you mentioned that you also co-founded a company, and I guess the field is kind of, is focused on fixing that. But my main question is, what convinced you that the biggest risk with projects isn’t necessarily technical or financial, that it’s more of a behavioral thing, especially when it comes to entrepreneurs? When did that kind of click for you?
Dr. Atif Ansar: So very early on in my research, I was looking at the construction of big dam projects. I’d grown up in Pakistan. My family and I had lived near, on a site for a big dam called Mangla, which is one of the two big dams Pakistan had built in the 1960s under a water treaty with India that the World Bank had financed. So my initial impression was that big dams were a good thing. They were sustainable power. It was, you know, basically once you’ve built the dam, you could supply clean energy for a very, very long period of time. You know, and as such, I had a very positive impression going into it. So my research when I began looking at it as an economist was to try and show the world that building more dams was a good idea for emerging markets as a way for them to have sustainable power. And what I found actually was a big, sort of changed my own mind and quite difficult for me emotionally to overcome, because what I found was the dams on average were doubling their budget and taking a lot longer to build than anybody had thought of, which was completely opposite to what I had originally imagined them to be, that they were going to have benefits that far exceeded the cost. And I found the opposite. And I realized that if I had not approached it with a kind of emotional neutrality, that as an academic you are almost drilled to approach it, I would have almost wanted to miss that data. And that’s when it clicked for me that overcoming one’s own psychological biases and only seeing what our brain wants to see is really difficult, and also requires a lot of courage to accept when facts disagree with your intuition. And then I began seeing those differences everywhere. So that’s when I got very interested in how behavior drives success or failure of projects rather than any technical method.
Erick Espinosa: And I noticed, it’s interesting that you say that because at some point through your research, you were able to kind of pinpoint that as a number, especially when it comes to mega projects. Do you want to share what that number is and I guess how that came about?
Dr. Atif Ansar: So my colleagues and I at Oxford built a very large data set of over 16,000 mega projects covering all kinds of assets, big dams that I mentioned, for example, but also road projects, rail projects, IT software programs, Olympics, space missions, you name it. So we sort of did very deep dives into specific verticals to try and understand what this phenomenon looked like. And we were interested in three or four core variables. So famously, there’s an iron triangle in project management of physical scope, cost and time that you might be looking to build. And then to that iron triangle, you can add dimensions like revenue or benefits, environmental impacts, social impacts, things like that. So we mainly looked at what do people promise in terms of the benefits of a particular project? What do people promise in terms of the costs? And what did they promise in terms of the timeline as the three core variables? And we tried to, of course, collect data on others. So in case of dam projects, I was interested in how much carbon did they try to abate and how much did they actually abate, so on and so forth. What we found was very interesting. So things like time, people say 10 years, they do 15 years. So there’s a big gap between hope and reality, and the gap is often adverse outcomes. So what people say versus what people do. Similarly, with budgets, like if people hope to achieve a particular project for a billion, say it becomes two billion dollars. And same thing with benefits, although it works the other way around, on rail projects, for example, people hope that, say, a thousand people will turn up for the rail line, but only 500 turn up. So cost overruns, time overruns and benefit shortfalls are common. They’re predictable, they’re systematic, and they happen to a vast number of projects. Now, if you look at individually in one of these dimensions, say nine out of 10 data centers are delivered late, for example. You could similarly say eight out of 10 dam projects are delivered over budget. So given these statistics of frequency of how many projects are delivered late, delivered over budget, so on and so forth, when you combine them, it basically means very, very, very few projects meet all three objectives. That is that they deliver on time, they deliver to the budget envelope, and they deliver on benefits. And that makes it a very interesting phenomenon to then look at, you know, the variance in the data, what explains the success versus failure of various projects.
Erick Espinosa: That’s what I think is so interesting about what you do, because the majority of pitches I get, and I did an episode a while back, I feel like a lot of the tools right now dedicated and saying, hey, we could address the delays in your big projects. So let’s say construction, right? Like there’s a lot of these AI projects that are now being like miscommunication, things are not organized properly, data’s all over the place, AI can come in and kind of fix that. But you’re also taking into account the mental aspect of all this. And what I think is really interesting is because I think on Foresight, there’s a quote here, and I want to say if I got this properly, no project should be delayed because the truth comes late. Which kind of relates to the ostrich effect, which you introduced me to, but was the company kind of built in that sense to kind of force people to look? And we’ll get into what the ostrich effect is exactly, but.
Dr. Atif Ansar: Absolutely. So look, you know, the thing with, so this idea of AI and data comes from Daniel Kahneman’s work. And Kahneman was a psychologist, you know, he worked, to the latter half of his life at Princeton University and won the Nobel Prize in Economics in 2002. He never really thought of himself as an economist, but his contributions from psychology were incredibly important to economics and in trying to correct a lot of economic theory that assumes rational human behavior, to think about ways in which human beings can behave systematically in irrational ways and how might that change, you know, economic hypotheses. And, you know, one of the interesting elements of that, he talks about in his 2011 book, which summarizes his life’s work, is Thinking, Fast and Slow. And the basic idea is human beings are largely automatic thinkers. You know, we react immediately. So if I were to say to you, you know, two plus two equals four emerges in your mind as an automatic response. It’s not a computation that you’re running in quite the same way a computer processes two plus two computationally. And because of that system one thinking, you know, we can recall, for example, months of the year, January, February, March, April, very fluently, but we’re not really, you know, processing that. Slow thinking is, if I was to ask you to multiply two very large numbers, if I give you two nine digit numbers to multiply, you can’t really do that in your head. You really have to sit down, write them out and multiply them doing slow thinking. Equally, if I ask you to put the months of the year in alphabetical order, suddenly that’s going to require use of a pen and paper or use of some kind of data.
Erick Espinosa: But the fingers at least on my hand will be like.
Dr. Atif Ansar: Exactly. So you can’t just kind of like work this out in your head because you might start with August and forget April or something like that. And, you know, jump straight to February and not think of December. Or, you know, there’s variety of ways in which people can mess these things up when we’re doing this in an automatic way. And that applies to optical illusions as well. So there’s a famous illusion, I’ve talked about this before around, you know, two lines and whether or not they look the same in terms of length or one looks longer than the other. The only way to overcome these things is often to use a foot ruler and measure and write down with your hand, you know, what is the estimated length of the two lines and whether or not they might be the same, you know, for example. And that’s the best use for AI and data is to really think of it as a way of correcting human biases, almost thinking that human beings’ natural way of looking at the world is rose-tinted and they almost need some kind of data-oriented, you know, evidence-based specs to correct their vision. And that’s kind of why I got into the world of AI and data. And that’s partly, you know, a way of overcoming biases like the ostrich effect that you just referred to.
Erick Espinosa: And can we get into the ostrich effect exactly? Like how would you define that specifically?
Dr. Atif Ansar: Absolutely. So Kahneman and Tversky back in the 1970s started a whole program in psychology which came to be known as heuristics and biases. And the question they asked was how do human beings form judgments under uncertainty? And some of these judgments are very simple. So how do you drive when it’s foggy? You know, how do you budget for a holiday? So not all of these judgments need to be very high stakes judgments. Some of them could be very simple judgments like how long is it gonna take me to go grab a coffee? So on and so forth. And in describing these mechanisms they found systematic evidence that people are biased, that we underestimate the time it takes us to go get a coffee, you know, or we tend not to slow down as much as we should on a foggy day. So we underestimate the risk of an accident on a foggy day, so on and so forth. That idea then basically led psychologists and researchers to look at variety of biases that human beings suffer from. And the ostrich effect is one such bias. And the basic formulation of that, you know, comes from a paper from Thomas Webb et al, is that when you’re faced with uncomfortable knowledge or risk, people have a tendency to stick their head in the sand like ostriches sometimes do.
Erick Espinosa: And so it’s a face— I didn’t understand why that term then, so that made more sense.
Dr. Atif Ansar: Exactly, because you find ostriches sticking their head in the sand when they’re like afraid of something, you know. And it’s kind of like a way of pretending that, you know, if I don’t see it, the risk doesn’t exist any longer. And that idea that psychologists found exists also. For example, if you’re overspending, people will avoid looking at their bank statements, you know, because it’s uncomfortable to do so. And there are multiple formulations of, you know, how the ostrich effect exhibits itself.
Erick Espinosa: I think that’s what a lot of people are gonna be doing at the end of August with their Europe trips. Everybody’s putting it on their credit cards and being like, I didn’t wanna look. That avoidance, right?
Dr. Atif Ansar: That is avoidance. It’s like, you’re not basically, you don’t wanna confront the uncomfortable knowledge. This idea also exists in social sciences. Steve Rayner was an anthropologist at Oxford. He had talked about this notion of uncomfortable knowledge as well. So, you know, basically inside organizations, also people have a preference to bury their heads instead of confront bad information that feels negative. Whether a project is running late or they’re gonna miss their revenue forecast. These are things that people tend not to wanna deal with, which is precisely what they need to do in order to move their project ahead.
Erick Espinosa: Okay, but the thing that kind of comes to mind for me, it’s not so much ignorance, but avoidance. So it’s you just being like, you’re choosing not to be aware, like instead of just not being aware, do you know what I mean? Of what’s going on?
Dr. Atif Ansar: That’s it. Well, a lot of these things are forms of self-deception. Unfortunately, this is a little bit of lying to oneself. So many of these biases, it’s not as if that we don’t know at all. Sometimes your brain just cannot do the job, like the optical illusion examples that I gave you. Other times it is a form of kind of lying to oneself and avoiding some information, because you don’t wanna confront it. It becomes a sort of an avoidance pattern rather than an ignorance pattern. My colleague, Alison Stewart, has done a little bit of work around willful ignorance in big projects.
Erick Espinosa: That was the word that came to mind for me just now.
Dr. Atif Ansar: It’s like people almost want to ignore and there’s almost a desire to ignore. So that phenomenon exists as well in social sciences.
Erick Espinosa: I think a lot of people could obviously relate to this from an individual perspective. But when we’re talking about like organizations in bigger projects, I mean, what does that really look like from the inside in terms of organizing these projects? Because people are leaning on dashboards, people are leaning on certain tools to organize their plans. But what you’re talking about is they may be choosing not to look at certain things in those tools when they may be realizing that it’s not going in the direction that they would have initially wanted.
Dr. Atif Ansar: So look, I think there’s in conventional wisdom, a view of the organization as some kind of a machine that’s emerged. That these big companies have these coherent minds and they have huge amounts of information and they really organize and really, really clever around how they operate. So as a result, people assume that big corporations are incredibly competent and incredibly rational and they don’t make mistakes or they don’t have biases. They’re able to, through organizational design, through governance processes, by able to process information, all of that, they’re able to make these very rational decisions. And you could sort of look at the way movies depict big companies as if they kind of know what’s going on. And as a result, they’re able to kind of deal with the external world in a very purposeful way. On the contrary, what management research shows them again and again, is that organizations are forms of messes. They are, there’s a famous paper from the 70s called the garbage can model, which is organizations almost represent garbage cans in terms of how messy they are. And in that, the question and the academic debate, which is also relevant in practice is, are organizations able to mitigate these human biases? So if an individual has a tendency to bury their head in face of risk, would a whole company do the same? And there’s quite a lot of debate. Some people say, no, organizations can override these individual biases by putting in place governance, et cetera. But contrary to that, what management research shows that organizations can be even more irrational than individual people and can behave in ways that are even more detrimental to their own goals than individuals. And we see that a lot in the project world. So the reason I’m interested as an academic in projects is that because they’re so focused, they represent a very interesting empirical device to study the broader machinations of an organization. And you can see that with the time pressure that these projects create, some of the both good and bad habits of organizations are magnified and amplified in that. And one such bad habit of organizations, particularly when they’re building projects, is inability to deal with the truth, which is often of a distasteful, of a bad news type of a situation. So a permit that’s running late, a customer that’s unhappy, a certain supplier that’s running late, a design that’s not been done on time, so on and so forth. People would rather waste their time debating the data, and excuse my French, bullshitting each other, than dealing with the truth head on and solving the problem. So I’ve observed this over and over again, people waste more time, you know, kind of essentially trying to pretend that things are okay, than acknowledging that a particular thing is not going okay and attacking that. And that is the problem of the ostrich effect writ large at projects. This then causes the kind of delays, the kind of failures we see in projects, where the delays are hidden in plain sight. You can see that a particular project is going to be late, but the management team is unable to attack that problem because of this bias.
Erick Espinosa: When you mentioned the governance thing, the first thing that comes to mind is, imagine you have an individual or a team assigned to that, something happens, it’s their job. And I feel like the first instinct maybe for most people would be pointing fingers, finding excuses, and everybody has their place in their role to bring the project to light. It does make sense in terms of, I mean, you see people that have been in their industry for a long amount of time, I’m sure they’ve seen this play out to some degree, whether or not they’ve done it, or somebody on their team or within their company has done it. But what makes me think right now in terms of the tools that we have available to try to address these things, how is that, what would you recommend people kind of do in terms of, to avoid this specifically, especially with the tools that exist? Is there anything that they could find useful in terms of kind of beating that ostrich effect that’s kind of innate?
Dr. Atif Ansar: One thing is to actually engage with the evidence in an emotionally mature way. In most companies, you can’t even find the data to establish whether projects are running on time or late. So the official narrative tends to be, we don’t have delays. So if you knock on the door of even very, very large companies today and you ask them, hey, how are your projects going? Their party line would be, everything’s done on time, everything’s fine. And poking through that and be like, hey, but really, like, do you have the historical data? Do you have the budgets available? Do you have the baseline schedule? Do you have the latest available? And you begin to realize that that claim is much hollower than you realize, right? And they’d realize themselves, that’s a more important thing. So it’s a little bit like a sales organization that had a revenue forecast that kept missing, but you can never find the revenue forecast gap. And so I think first step really, Erick, is just acknowledging and understanding the baseline data. You know, what did you say? What actually happened? And doing that without guilt, without blame, you know, just in a very emotionally neutral way.
Erick Espinosa: I think that’s very key in terms of the wording that you use. Because I think most people would be like, you’re micromanaging me. You’re going into what I’m doing and you’re not letting me do the work that I need to do. And I think the wording is very important in terms of the way that you approach it, because some people could feel like you’re automatically accusing them of something when you’re all kind of, you know, in this project together.
Dr. Atif Ansar: That’s it. You know, it’s a little bit like fitness. If somebody is trying to become an athlete and they need to track their performance of how well, for example, somebody is trying to become a power lifter and their goal is to achieve, for example, like a 200 kilo deadlift. Like they need to understand how much can they safely deadlift today. And if they refuse to look at the number of plates they’ve put on their bar and say they can deadlift 60 kilos a day, there’s no way they can make progress to 200 kilos without baselining that data for themselves and knowing where they are right now and what sort of, you know, regime they need to put themselves through. So, you know, people do find that very difficult to establish those baselines. And instead of, you know, and if somebody was to do that for them and to say, hey, you can only lift 60 kilos right now. Your goal is 200 kilos. This is what it would take to get there. You know, the ostrich effect would be to bury your head in the sand and refuse to even believe that that 60 kilo number is correct. Like, no, no, no, it was not 60 kilos. It was really 120 kilos. That would be sort of, or it was already 180 kilos. It’s sort of like thing that people can have. So it’s kind of going back to your question. The solution to this is first of all, they have to acknowledge the problem. So, and find the data. And if you can’t find the data, that in itself is a telltale sign. Many organizations suffer from the fact that they can’t even locate that basic budget information. And I think it is difficult for big companies to set aside feelings of embarrassment around this. So, and I want to reinforce the point around, it is really, really important to adopt an emotionally neutral stance. Not like either get angry about it or sad about it. Just kind of like find the golden mean around what is the information we can find today? And how has it gone in the past as a way of baselining and setting an objective for where we can get to in the future? Will all your projects be delivered on time immediately after you have knowledge of some projects running late or many projects running late? No, but even understanding, look, we want at least half our projects to be delivered on time by next year. What does that mean? Does that mean that we need to improve our estimation? Does that mean we need to improve our contracting methodology? Just sort of, you need a period of self-reflection with data, a little bit like the slow paper exercise that I was talking about. Again, a lot of organizations avoid that. They sort of say, oh, but this is creating too much change. I don’t want to do the hard work. That’s ultimately hurting their own outcomes. And third, and in many ways, most important is you’ve got to act. So once you’ve made up your, once you’ve seen the data and you’ve got to act as an organization, I think here, the third piece is where many companies let themselves down. Action becomes incredibly diffuse. The CEO delegates to CEO minus one, CEO minus one delegates to the next level down. By the time the 18th person is involved, they either don’t have the context or the strategic vision to really drive that change. The action becomes incredibly tactical and becomes completely diffused. And then there’s just no way the whole company is going to reorient itself.
Erick Espinosa: When I think about this in terms of on a smaller scale, because you work with a lot of bigger mega projects, a lot of these things apply to people that work as small entrepreneurs. Let’s say they have a team of 50 or like smaller, but they maybe outsource certain tasks and projects to third parties. How do you kind of keep up with that in terms of the ostrich effect when you’re kind of working with a company that’s assigned to a project, but you want to make sure everything’s going properly there, but don’t necessarily have all that kind of relationship or direct access as you would internally?
Dr. Atif Ansar: So one thing is that you start questioning greens and start supporting reds, which is the opposite of what people normally do. So one is know that in organizational environments, numbers that are hitting their forecasts are usually quite suspect. So if you’re meeting your revenue forecast year in or like month in, month out, like that will look green on the dashboard, query that very, very extensively.
Erick Espinosa: That’s good food for thought. That’s very good food for thought for sure.
Dr. Atif Ansar: So a startup entrepreneur has to know that themselves and people around them heavily suffer from optimism bias and potentially more than anybody else because they’re startup entrepreneurs. So they have to work even harder to overcome the ostrich effect and the optimism bias. And they do that by, if you’re being told that everything’s hunky-dory, question that heavily and be skeptical. Without necessarily, and again, without emotion, maybe that’s the heart piece actually, which is you don’t have to shout at people, but like really test what was our genuine forecast. And more importantly, as an entrepreneur, you can’t lie to yourself. So you really can’t go back and retrospectively revise your numbers to say, oh, but we didn’t intend to sell X thousand this month. We only intended to sell that. So which is the tendency that people dilute their own goals. So to take a fitness example again, if you wanted to achieve a certain number and you miss it that quarter, you can’t go back and say, well, I didn’t actually mean to achieve that number. I meant to achieve the number that you actually achieved.
Erick Espinosa: And in terms of the mega projects, because I know, I guess the most recent types of mega projects, the most people are kind of familiar with right now are data centers. And that’s also something that you address, teams that you work with. And I have a number here. It says analysts estimate that 30 to 50% of large scale data center capacity planned for 2026 will be delayed. And one checker found that about 6% of projects announced in 2025 were rated high likelihood to actually execute. What is your conversations, I guess, with these data centers, just very similar to all the mega projects that have existed before. Like what are those kinds of conversations look like with those teams nowadays to make it more realistic?
Dr. Atif Ansar: Data center industry is an interesting one, Erick, because it represents probably some of the most sophisticated construction teams. So data centers now account for nearly 10% of global construction volume.
