AI Maturity as a Double-Edged Capability: Planning Fallacy and Firm Performance in Technology Startups
This study reveals that while AI maturity generally boosts technology startup performance, it paradoxically exacerbates the negative impact of planning fallacy, acting as a double-edged sword that requires disciplined planning to realize its full potential.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine you are the captain of a rocket ship trying to reach a new planet. You have two main things to worry about: the quality of your engine and the accuracy of your map. In the world of business science, this is like studying how Artificial Intelligence (AI) works in new companies. AI is like a super-powered engine that can process information, automate tasks, and help make decisions faster. But having a powerful engine doesn't guarantee you'll get to your destination; you also need a good map.
The paper looks at a common human mistake called the planning fallacy. This is when people are overly optimistic. They think a project will take less time, cost less money, and have fewer risks than it actually will. It's like guessing a road trip will take four hours because you've never hit traffic, ignoring the fact that you usually do. The researchers also look at AI maturity, which isn't just about having one cool robot. It's about how deeply a company has built AI into its daily habits, rules, and team structure. The big question is: If a company has a super-advanced AI engine, does that fix the captain's bad map? Or does a powerful engine actually make the crash worse if the map is wrong? This matters because many new tech startups are betting their future on AI, and they need to know if their optimism is helping or hurting them.
The Double-Edged Sword of Smart Machines
This study dives into the lives of 165 technology startups in Tehran, Iran, to see how the mix of "bad planning" and "smart AI" affects how well a company does. The researchers surveyed the bosses and founders of these companies, asking them about their tendency to underestimate costs and overestimate success, as well as how mature their AI systems were. They used a special math tool called Partial Least Squares Structural Equation Modeling (think of it as a high-tech calculator that finds hidden patterns in survey answers) to crunch the numbers.
Here is what they found, and it's a bit of a twist on what many people might expect.
The Bad News: Optimism is Still Dangerous
First, the study confirmed what many experts suspected: planning fallacy is bad for business. When founders consistently guess that projects will be cheaper and faster than they really are, the company's performance suffers. It's like trying to bake a cake with a recipe that says you only need one egg when you actually need three; the result is a mess. The data showed a clear negative link: the more the founders underestimated risks and costs, the worse the company performed.
The Good News: AI Usually Helps
On its own, having a mature AI system is a good thing. The study found that companies with well-integrated, "mature" AI capabilities tended to perform better. It's like having a high-performance engine; generally, it makes the car go faster and smoother.
The Twist: The "Double-Edged" Effect
Here is where it gets interesting. The researchers discovered that AI maturity acts like a double-edged sword. While AI is generally good, it actually makes the damage caused by bad planning worse.
Imagine you are driving a car. If you have a slow, old car and you guess the trip will take 30 minutes when it actually takes an hour, you might just be a little late. You can probably handle it. But if you are driving a rocket-powered car that goes 200 miles per hour, and you make that same wrong guess, you aren't just late; you might crash into a mountain.
The study found that for startups with high AI maturity, the negative impact of planning fallacy was much stronger. In other words, the more advanced and integrated a company's AI was, the more damage their overly optimistic plans caused. The AI didn't fix the bad map; instead, it helped the company execute the bad plan faster and more efficiently, leading to bigger problems.
What the Study Rules Out
The paper explicitly argues against the idea that AI is a magic cure-all. It suggests that simply buying or building advanced AI does not automatically stop managers from being overconfident. The AI doesn't "think" for itself to correct the boss's mistakes; it just follows the instructions given to it. If the instructions are based on a fantasy, a super-smart AI will just help you build that fantasy faster.
How Sure Are We?
The researchers are quite confident in these associations based on their data from 165 startups. They found that the link between bad planning and poor performance was statistically significant, and the "worsening" effect of AI was also a clear, measurable result. However, because the study was done at a single point in time (a snapshot) and relied on the bosses' own opinions about their past projects, the authors suggest these are strong associations rather than absolute proof of cause-and-effect. They also noted that the "worsening" effect, while real, was a "small" statistical effect size, meaning it's a subtle but important nuance in the complex world of business.
The Takeaway
The main lesson for any tech startup is simple: Don't let your AI engine outpace your map. Having a sophisticated AI system is a huge advantage, but it only works if the people running it are realistic about time, money, and risk. If you have a powerful engine but a terrible map, the more powerful the engine is, the harder you will hit the wall. The study suggests that to get the most out of AI, companies need to pair their technology with disciplined, realistic planning routines that challenge their own optimism.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.