Chesterton's Fence: Don't Remove What You Don't Understand

A new head of growth joins, opens the campaign dashboard, sees a "boring" retargeting flow nobody can explain, and kills it in week one. Three months later, the pipeline number everyone quietly relied on is gone, and nobody remembers why the flow existed until it is too late.

Quick Summary

• Chesterton's Fence: before removing a rule, process, or campaign, first understand why someone put it there.
• The failure mode is common in marketing: new leaders or new hires cut "legacy" things they find embarrassing or confusing, without asking who built them and why.
• The fix is not "never change anything." It is "understand first, then decide," on a timeline proportional to the risk.
• Direct mail still delivers a 161% ROI to house lists precisely because it survived years of "let's cut the boring channel" meetings.
• Two guardrails keep the principle from becoming an excuse for inertia: a time-boxed investigation, and an explicit bias for reversible changes.

The Principle and Its Origin

The idea comes from G.K. Chesterton's 1929 book The Thing. He described a fence built across a road for no reason a passerby could see, and the reformer who wants it gone immediately.

Chesterton's answer, in his own words, is the whole principle: If you do not see the use of it, I certainly will not let you clear it away. Go away and think. Then, when you can come back and tell me that you do see the use of it, I may allow you to destroy it.

Notice what he is not saying. He is not saying the fence must stay forever, and he is not saying old things are automatically wise. He is saying: understanding comes before removal, not after.

That ordering is the entire skill. Skip it, and you inherit consequences nobody warned you about.

The Marketing Failure Mode

Marketing accumulates fences fast: approval workflows, legal-mandated disclaimers, a "boring" email cadence, a channel nobody can explain in a Slack thread. Each one looks like friction to a fresh set of eyes.

New CMOs and new hires are especially exposed. They arrive with a mandate to "modernize," a natural urge to prove impact quickly, and zero institutional memory of why anything exists. The two-year-old approval step that looks like bureaucracy might be the exact control that kept a past regulatory fine from happening twice.

The 2010 Gap logo redesign is the visible, public version of this pattern: a legacy mark, quietly carrying decades of recognition equity nobody had bothered to measure, replaced in a weekend. Public backlash forced Gap to revert within six days. The fence was brand memory. Nobody asked why it was there before tearing it down.

The quieter version happens weekly, never in the trade press. A "boring" retargeting flow, an "outdated" direct-mail segment, a "redundant" legal review step, cut because nobody currently on the team remembers commissioning it. The channel or process was carrying weight invisible to whoever inherited it.

That distinction, missing information versus missing value, is the one reformers skip. Fix the information gap first.

A Framework: Understand Before You Remove

The framework has four steps, and it is designed to be fast, not to stall every decision for a quarter.

Find the builder. Someone approved that process or launched that campaign. A five-minute Slack search or a coffee chat with the person who has been there longest often answers 80% of the question.

Reconstruct the data. Pull the performance history before you pull the plug. A "stale" email cadence that quietly converts 12% of monthly pipeline is not stale, it is unglamorous and load-bearing.

Check for external constraints. Legal, compliance, and partner agreements are the most dangerous fences to misread, because their absence shows up as a fine or a lawsuit, not a dip in a dashboard.

If you still cannot find the reason, do not remove, test. Turn the fence off for one segment, one week, one market. Measure what breaks before you scale the removal company-wide.

Each step costs hours. Skipping all four can cost a quarter of pipeline, or a compliance incident that costs a lot more than a quarter.

When Not to Over-Apply It

Chesterton's Fence is not a license for "we've always done it this way." That is the opposite failure, and it is just as expensive.

The principle applies to fences you do not understand. It does not apply to fences you understand perfectly well and know are broken. If everyone on the team can already explain exactly why a process exists and agrees the reason no longer holds, you do not need an investigation, you need a decision.

The other guardrail is reversibility. A one-week test of pausing a "confusing" campaign segment is cheap to undo. A full rebrand or a deleted compliance workflow is not. Match your investigation depth to how hard the change is to reverse, not to how annoying the fence looks today.

Understanding first does not mean waiting forever. It means matching the size of your inquiry to the size of the risk, then moving.

Key Takeaways

• Chesterton's Fence: understand why a rule, process, or campaign exists before you remove it, not after.
• New marketing leaders are the highest-risk group for this failure, because they have the least institutional memory and the most incentive to "modernize" fast.
• Legacy or "boring" channels like direct mail often carry real, measurable ROI that a quick glance misses.
• Run the four-step framework: find the builder, reconstruct the data, check external constraints, test small if you still don't know.
• The principle is not "never change anything." If you already understand why a fence exists and why that reason no longer holds, remove it and move on.