There is a comfortable belief inside most property companies that the reason a hard decision has not been made yet is that the data is not quite complete. One more report. One more month of numbers. One more cut of the occupancy trend, the collections aging, the maintenance cost per unit, and then the right call will become obvious and the decision will make itself. So the analysis gets commissioned, and the decision waits.

The belief is mostly wrong, and often backwards. Past a fairly early point, more data does not make a decision clearer or a decision-maker bolder. It does the opposite. It delays the decision, inflates confidence without improving judgment, and quietly supplies everyone involved with permission to keep not deciding. The problem the extra data was supposed to solve, indecision, is frequently made worse by the very thing prescribed to cure it.

More information mostly buys confidence, not accuracy

The most uncomfortable finding in this area is old, well-replicated, and almost never acted on. Additional information reliably makes people more confident in their judgments without making those judgments more correct.

The classic demonstration is a 1965 study by the psychologist Stuart Oskamp. He gave clinicians a case study divided into four parts and asked them to answer the same set of questions after each part, so their information about the person grew in stages while the questions stayed fixed. As they received more information, their confidence in their answers rose steadily, but their accuracy did not improve to match, so the gap between how right they felt and how right they were widened with every additional page. By the end, the large majority were overconfident. More data had made them feel more certain and left them no more correct.

For a decision-maker this is the crux of the matter. The extra report does something, it raises your confidence, and confidence feels like decisiveness. But feeling more sure is not the same as being more right, and a decision made from inflated confidence is not a better decision, only a more comfortable one. A great deal of the data gathered before a decision is bought precisely for that comfort, and it is worth being honest that comfort, not accuracy, is often what it delivers.

More options can produce fewer decisions, not better ones

There is a second mechanism, and it runs directly against the intuition that richer information sharpens choice. Beyond a point, more to consider makes people less able to decide at all.

This is the phenomenon of choice overload, or overchoice, the finding that a larger set of options can impair decision-making rather than improve it, a term that dates to Alvin Toffler in 1970 and was given its best-known demonstration by Iyengar and Lepper in 2000. In their study, a supermarket tasting table offered either six varieties of jam or twenty-four. The larger display drew more interest, but far fewer of the people who stopped at it actually bought anything, a roughly tenfold collapse in the rate of action compared with the smaller display. More options attracted attention and then suppressed decision.

The honest footnote is that this specific effect has a mixed replication record, later meta-analysis found choice overload is real but context-dependent, strongest when the options are complex and the chooser's preferences are uncertain, which happens to describe most consequential business decisions rather well. Applied to data rather than jam, the mechanism is the same. Every additional variable, cut, and scenario is another option the mind has to weigh, and past a threshold the weighing does not converge on a clearer answer, it fans out into more considerations, more caveats, and more reasons the decision is not quite ready. The analysis meant to enable the decision becomes the reason it keeps not happening.

The deeper problem: data as permission not to decide

Underneath both mechanisms is something more organizational than psychological, and it is the real reason data accumulates in front of hard decisions: more data is the most respectable available excuse for not deciding.

Every difficult decision carries the risk of being wrong, and being wrong is personally costly in a way that is visible and attributable. Commissioning more analysis is the perfect hedge against that risk. It looks diligent, nobody can criticize a leader for wanting to be thorough, and it postpones the moment of commitment without anyone having to admit that commitment is what is being avoided. "Let's get more data" is the most defensible sentence in any meeting, which is exactly what makes it so dangerous: it converts avoidance into apparent rigor, and it can do so indefinitely, because there is always another cut of the numbers to request.

This is how a decision that should take a week takes a quarter, with each round of analysis spawning the questions that justify the next round. And when the decision is finally made, more data has quietly done one more unhelpful thing: it has diffused the accountability for it. A choice made after ten reports, three models, and a steering committee is a choice no single person really owns, which feels safer to everyone and is worse for the organization, because a decision nobody owns is a decision nobody is truly accountable for getting right. The data did not sharpen the decision. It spread the responsibility for it thin enough that no one has to carry it.

The honest part

Several qualifications matter, because the opposite error, deciding blind, is just as real and this argument can be misused to justify it.

Some data is genuinely necessary, and this is not a case for winging it. There is a real threshold below which you do not have enough to decide responsibly, and a leader who prides themselves on going with their gut while ignoring the small number of facts that actually bear on the decision is not being decisive, they are being reckless. The argument is about what happens past the point of sufficiency, not below it. The skill is knowing where that point is, and it is usually reached far earlier than the appetite for more analysis suggests.

It is also true that some decisions genuinely warrant deep analysis, and treating all deliberation as avoidance is its own mistake. A large, irreversible, bet-the-company decision deserves more data and more time than a small reversible one, and the discipline is proportioning the analysis to the stakes and the reversibility, not minimizing it everywhere. The problem is not analysis. It is analysis disconnected from any threshold at which it will stop and a decision will be made.

And more data really can improve a decision, up to the point where it starts serving confidence and delay instead of judgment. The claim is not that information is the enemy. It is that information has a job to do in a decision, and once it has done that job, additional information stops helping and starts hurting, and most organizations have no mechanism for noticing when that line has been crossed.

Decide from a threshold, not from exhaustion

The practical discipline is to change what triggers the decision. Most organizations decide when they run out of appetite for more analysis, which is a moving target that recedes as the decision gets harder. The alternative is to decide from a pre-set threshold of sufficiency, defined before the analysis begins.

Before commissioning the analysis, answer three questions. First, what specifically would this data change? If you cannot name a decision that would go differently depending on what the data shows, the data is not decision-relevant, it is comfort, and you can skip it. This is the same test applied to reporting in general in the reports nobody reads are about to multiply, asked here about a single decision rather than a standing report: information that changes no action is not informing anything.

Second, what is the smallest set of facts that would let a reasonable person decide? Define that up front, gather exactly that, and treat it as the threshold. When you have it, you decide, not when you have run out of things to ask for.

Third, who owns this decision? Name the single person accountable before the analysis starts, so that the data supports their judgment rather than replacing it, and so the decision cannot dissolve into a committee where more reports substitute for a responsible owner. This guards against the quiet way abundant data converts a decision into a number everyone can point to instead of a call someone has to make, a cousin of the distortion examined in the version of the numbers you started to believe, where the presented figure gradually replaces the underlying judgment.

There is one question that reveals whether more data is serving you or stalling you. If the next report came back exactly as you expect, would you make a different decision than you would today? If the answer is no, you already have enough to decide, and the report you are waiting for is not analysis, it is delay wearing the costume of diligence. The decisive organizations are not the ones with the most data. They are the ones that know when they have enough and then actually decide, while the others are still requesting one more cut of numbers that was never going to change the answer.

FAQs

Q1. Doesn't more data lead to better decisions?
Up to a point, yes, and below a minimum threshold you genuinely need more. But past the point of sufficiency, additional data mostly raises confidence without raising accuracy, adds options that make deciding harder, and supplies cover for postponing the decision. The relationship between data and decision quality is not a straight line upward; it rises, plateaus, and then bends down as more information starts serving comfort and delay rather than judgment.

Q2. What does the Oskamp study actually show?
In a 1965 experiment, clinicians answered fixed questions about a case study as they received it in stages. With each additional part, their confidence in their answers rose, but their accuracy did not improve to match, so they became steadily more overconfident. The lesson for decision-makers is that extra information reliably makes you feel more certain without necessarily making you more correct, and feeling certain is easily mistaken for being right.

Q3. How can more information make us less able to decide?
Through choice overload: beyond a threshold, a larger set of options or considerations impairs decision-making rather than aiding it. The well-known jam study found that a display of twenty-four varieties drew more interest but produced far fewer purchases than a display of six. Applied to data, every extra variable and scenario is one more thing to weigh, and past a point the weighing fans out into more caveats rather than converging on a clearer answer.

Q4. Isn't the jam study contested?
Yes, and it is fair to say so. The specific effect has a mixed replication record, and later meta-analysis found choice overload is real but depends on context, appearing most strongly when options are complex and the chooser's preferences are uncertain. That caveat matters, but it also happens to describe most significant business decisions, which are exactly the complex, high-uncertainty situations where the effect is most likely to bite.

Q5. What do you mean by data as "permission not to decide"?
Commissioning more analysis is the most defensible way to avoid a hard decision. It looks diligent, no one can fault a leader for being thorough, and it postpones commitment without anyone admitting avoidance is the goal. Because there is always another cut of the numbers to request, this can continue indefinitely, turning a week-long decision into a quarter-long one while each round of analysis generates the questions that justify the next.

Q6. How does more data diffuse accountability?
A decision made after many reports, models, and committee reviews becomes a decision no single person really owns. That feels safer for everyone involved, because responsibility is spread across the analysis and the group, but it is worse for the organization, since a decision nobody owns is one nobody is truly accountable for getting right. Abundant data can quietly convert a personal judgment call into a diffuse outcome that no individual has to stand behind.

Q7. Aren't you just arguing for gut-feel decisions?
No. Deciding blind is as real an error as over-analyzing, and there is a genuine threshold below which you lack enough to decide responsibly. The argument concerns what happens past sufficiency, not below it, and large or irreversible decisions genuinely warrant more analysis than small reversible ones. The point is to proportion analysis to the stakes and stop at a defined threshold, not to minimize information everywhere or to celebrate ignoring it.

Q8. What is the single most useful question to ask?
Ask whether you would decide differently if the next report came back exactly as you expect. If the answer is no, you already have enough to decide, and the report is delay rather than analysis. Pairing that with a pre-set threshold of sufficiency, defined before the analysis begins, and a named owner for the decision, keeps data in the role of informing a judgment rather than replacing or postponing it.