Subtractive Neglect Bias: The Hidden Reason Your Processes Keep Getting More Complicated

Walk into almost any operations review and you'll hear some version of the same conversation. Something went wrong, so the group starts asking what to add. Another approval. Another checklist item. Another field on the form. Another dashboard tab. Nobody in the room ever asks the opposite question: what's already in this process that shouldn't be there.

That instinct has a name. Researchers call it subtractive neglect bias, sometimes shortened to subtraction neglect or additive bias. It describes a simple, well-documented pattern: when people try to improve something, they overwhelmingly reach for addition, even when removing something would work better, cost less, and take less time.

The Lego Study That Explains a Lot

In 2021, a research team from the University of Virginia (Gabrielle Adams, Benjamin Converse, Andrew Hales, and Leidy Klotz) published a study in Nature that made this bias hard to ignore. In one experiment, participants were shown a Lego structure that needed to be sturdy enough to support a brick without collapsing. They were given a dollar for the task and could buy extra Lego pieces for ten cents each, or they could simply remove a piece that was already there, for free.

About 60% of participants added bricks. The cheaper, faster, more effective fix — pulling out the one piece that was actually causing the instability — occurred to a minority of people. Klotz later expanded on this research in his book Subtract: The Untapped Science of Less, arguing that this pattern shows up everywhere from city planning to product design to how we run meetings.

The business version of this is painfully familiar to anyone who has sat through a process improvement discussion.

Why We Default to Adding

Addition is visible. If you add a step, a field, a check, or a report, you can point to it. You did something. Subtraction is quieter. It requires you to look closely at a process that already exists, figure out which pieces are load-bearing and which aren't, and then make a call to remove something someone built on purpose, at some point, for a reason that may or may not still apply.

That last part matters. Removing a step feels riskier than adding one, even when the data says otherwise. Nobody gets blamed for adding an approval layer. Someone might get blamed for removing one, even if it was never doing anything.

Where It Shows Up in Your Business

Client onboarding. A client slips through a gap during setup. The response is almost always to add a step: another sign-off, another review meeting, another checklist item. Rarely does anyone go back and ask why the steps that already existed didn't catch the problem. A year later, onboarding has grown from six steps to nine, takes twice as long, and the original gap may still be open.

Approval chains. One bad decision gets through, so leadership adds a layer of sign-off. Repeat that cycle a few times over a few years and you end up with an approval chain nobody fully understands, where most of the signers are rubber-stamping something three other people already reviewed.

Dashboards and reporting. Every dashboard starts lean. Then a stakeholder asks for one more metric. Then another. Eighteen months later, the dashboard has thirty widgets, nobody remembers why half of them are there, and the report that used to get read every week now gets skimmed once a month, if that. The fix is almost never a better dashboard. It's cutting back to the five to seven numbers that actually drive a decision.

Account reconciliation. This one hits close to home for anyone reconciling data across a practice management system, QuickBooks, and a bank feed. A discrepancy shows up, so someone adds a manual cross-check to the monthly process. The root cause gets fixed a few months later, but the manual check stays, because removing it was never on anyone's task list. Multiply that by a few years and a few discrepancies, and a bookkeeping process that should take two hours takes six, with half the time spent on checks that stopped mattering long ago.

Meetings. A status update gets missed, so a new recurring meeting gets created to make sure it doesn't happen again. Nobody asks whether an existing meeting could have covered it, or whether the real issue was that the existing meeting wasn't being used well.

Subtractive Neglect Bias, in Practice

Client onboarding process — two ways to respond when steps get missed

The default move is almost always add. The better move is usually to look at what's already there.

The Fix: Subtract First

None of this means addition is always wrong. Sometimes a process is genuinely missing a step, and adding one is the right call. The problem is that most teams never ask the subtraction question at all. Addition is the default, not a decision.

A simple habit fixes most of this: before approving any change meant to fix a broken process, ask what could be removed instead of what should be added. When something breaks, look first at the steps closest to the failure and ask which ones are actually doing anything. If a step, field, approval, or report can't clearly earn its place, it's a candidate for removal, whether or not it's connected to the current problem.

This is also a useful lens for anyone building templates and standard operating procedures meant to be reused across clients or projects. Every time a template gets copied and adjusted, it tends to pick up small additions from whoever used it last. Left unchecked, a template built to save time slowly turns into one that adds it back.

The businesses that stay fast and easy to run aren't the ones that never add anything. They're the ones that ask the subtraction question as often as the addition one, and actually act on the answer.

Source: Adams, G. S., Converse, B. A., Hales, A. H., & Klotz, L. E. (2021). People systematically overlook subtractive changes. Nature, 592, 258–261.