# The Library That Never Threw Away a Book.

The building had been a library since 1923.

Seven floors. Roughly 340,000 books across every subject the original founders thought worth preserving. The collection had grown every decade since — donations accepted without condition, purchases made by librarians who retired and were replaced by others who continued the same tradition, acquisitions logged into systems that tracked what came in but never what was actually being read.

By 2019, the library had a problem it had spent almost a century creating.

The shelves were full. Not full in the way a thriving library is full — books checked out, returned, worn at the spine from use. Full in the way a storeroom is full. Rows of books that had not been opened in fifteen years. Multiple copies of titles nobody had requested since the 1980s. Entire sections serving subject areas that no living borrower had ever enquired about.

A facilities report that year found that 34% of the collection had not been accessed in any recorded way for more than a decade.

The librarians knew this. They had known it for years.

When a consultant suggested a systematic cull — removing books that had not been borrowed in ten years, freeing shelf space for titles people actually wanted, making the library navigable again — the response was immediate and unanimous.

They could not throw away books.

Not because they had evaluated each one and determined it was worth keeping.

Because it was theirs.

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/7d912f05-16fa-4399-886f-49d724e4e3dd.png align="center")

* * *

### The Mechanism Underneath Every List That Is Too Large

In 1980, an economist named Richard Thaler published a paper in the Journal of Economic Behavior and Organization that named something every person who has ever owned anything already knew but had never examined properly.

He called it the endowment effect.

Thaler's finding was precise: people consistently assign more value to things they own than to equivalent things they do not own — not because they have evaluated the owned thing more carefully, but simply because ownership itself changes the psychological relationship to the object. The moment something becomes yours, giving it up registers not as a neutral transaction but as a loss. And as Thaler's later work with Daniel Kahneman and Jack Knetsch confirmed, losses hurt significantly more than equivalent gains feel good.

The laboratory demonstration of this effect is almost embarrassing in its simplicity. Participants were randomly given a coffee mug. They were then asked the minimum price for which they would sell it. A separate group, who had not been given the mug, were asked the maximum they would pay to buy one. Same mug. Same day. Same participants in every other meaningful way.

The owners demanded an average of $5.25 to give it up.

The buyers would pay an average of $2.25 to acquire it.

The mug had not changed. Ownership had.

Now look at the email list sitting in your marketing platform right now.

How many subscribers on it have not opened a single email in eighteen months?

How many have not opened anything in twenty-four months?

Now ask the harder question: when did your team last have a serious conversation about removing them?

If the answer is rarely, or never, or "we discussed it but decided to wait" — you are looking at the endowment effect operating on a subscriber list. The list is yours. It was acquired at cost, through campaigns and lead forms and years of effort. Giving any part of it up registers not as optimization but as loss — even when the dormant portion of it is actively destroying the value of the active portion.

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/90d01ebf-fe31-4145-984a-8049e6e3925d.png align="center")

* * *

### What the Dormant Subscriber Is Actually Doing to the List Around It

This is the part that makes the endowment effect genuinely dangerous in email marketing, as distinct from merely costly in the way it is in other contexts.

In most domains, overvaluing something you own is an error that costs you the difference between what you think it is worth and what it actually is. You hold a stock too long. You price a house too high. You keep furniture in storage that costs more per year to store than the furniture is worth.

The cost is bounded. You lose the value of the thing.

In email, the cost is not bounded. The dormant subscriber does not simply sit there, worthless. They actively damage the value of every other subscriber on the list.

The mechanism works like this. When you send to a list that contains a significant proportion of dormant, disengaged addresses, the receiving mail servers — Gmail, Outlook, Yahoo — observe that a meaningful percentage of the people you are sending to are not engaging. They are not opening. They are not clicking. In some cases they have moved to a new address and the old one is now a spam trap. In some cases they are human beings who have simply trained their inbox filter to route your emails directly to promotions or spam, which Gmail tracks as a negative engagement signal regardless of whether you can see it in your open rate.

These signals accumulate into your domain's reputation score. And the reputation score does not differentiate between the dormant subscriber and the genuinely engaged one when it is deciding where to route your next campaign. The damage done by the 40% who never open flows directly into the inbox placement of the 60% who do — or used to.

The library that keeps every book regardless of whether it has been read in fifteen years is not merely wasting shelf space. It is making it progressively harder to find the books anyone actually wants.

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/b56fbf60-a4fd-4b93-98f9-bf4ca386ec64.png align="center")

* * *

### The Four Ways List Hoarding Shows Up in the Meeting Room

These four symptoms appear in a specific sequence. Recognise your programme in the first and the second, third, and fourth are already developing.

* * *

**"Our Open Rates Have Been Falling for Two Years and We Can't Explain It"**

The most common symptom of the endowment effect applied to a subscriber list. Open rates decline gradually. The explanation offered in meetings rotates through subject lines, send times, content quality, and seasonality — each one tried, each one producing a temporary recovery that fades, none of them addressing the underlying cause.

The cause is usually visible in the list itself. The proportion of subscribers who have not opened anything in twelve or more months has been growing quietly, pulling the aggregate open rate down not because the engaged subscribers are becoming less engaged but because the dormant ones are accumulating weight in the denominator.

*The librarians knew the books weren't being borrowed. The problem was visible for years before anyone acted.*

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/f35c13e0-bb6d-4a05-a3b6-370b3da13412.png align="center")

* * *

**"We're Getting Gmail Promotions Tab Placement and We Don't Know Why"**

Promotions tab routing is Gmail's way of saying: our signals suggest this sender does not reliably produce content the recipient wants to see in their primary inbox. The signal that most commonly produces this routing is aggregate engagement data across the sender's full list — including the dormant portion.

A brand sending to a list that is 40% dormant is giving Gmail 40% of its engagement data from people who do not engage. Gmail's routing decision reflects this. The 60% who would have happily received the email in their primary inbox now find it in Promotions instead — because the 40% the brand refused to remove has spoken on their behalf.

*The valuable books are still there. They are just buried behind the ones nobody wants.*

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/9001d69b-0e5b-41b8-88f6-96351fbf7f44.png align="center")

* * *

**"We Cleaned the List Once and Our Metrics Got Worse Immediately"**

This symptom is the endowment effect's most insidious contribution: it makes the cure look like the disease.

A brand removes a segment of dormant subscribers. List size drops visibly — say, from 800,000 to 480,000. The next campaign goes to 480,000 people instead of 800,000. The absolute number of opens drops. Someone in the room says: "cleaning the list made our numbers worse."

What actually happened: the relative engagement rate almost certainly improved — the opens as a percentage of sends went up, because the denominator shrank faster than the numerator. The domain reputation signal improved. But the absolute number went down, and absolute numbers are what most dashboards show by default.

The brain that overvalues what it owns reads the smaller list as a loss even when the smaller list is performing better. The endowment effect produces the fear of cleaning, and then — when cleaning happens — produces the misreading of the result that justifies stopping.

*The library that culled a section of its stacks saw open floor space for the first time in decades. The head librarian filed a report noting that the total book count had declined.*

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/9700204a-bf8b-46ea-95bb-c73c4f93a43f.png align="center")

* * *

**"We're Seeing Deliverability Issues but Our Complaint Rate Looks Fine"**

The most technically advanced symptom, and the one that catches the most experienced email teams off guard.

Complaint rate — the percentage of recipients who mark your email as spam — is a lagging indicator. By the time it registers in Postmaster Tools, the reputation damage has been accumulating for weeks, driven not by active complaints but by passive non-engagement from dormant subscribers.

Gmail's quality signal in 2026 runs on both complaint rate and engagement rate. A list with low complaints but very low engagement — which is exactly what a list with a large dormant segment produces — still triggers routing changes and eventual inbox placement degradation.

The complaint rate looks fine. The inbox placement is falling.

The mug still looks like a mug. It just isn't worth $5.25 anymore.

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/ce8416c0-d2df-4a50-9eb6-e9001d090353.png align="center")

* * *

### What Thaler's Research Actually Prescribes

The endowment effect is not a character flaw. Thaler was explicit about this — it is a documented, predictable feature of human psychology operating on a perfectly rational system. Losses feel larger than equivalent gains because the brain's loss-aversion circuitry evolved to protect against exactly the kind of irreversible depletion that made losing things dangerous for most of human history.

The fix is not to override the feeling. It is to change the reference point.

Thaler's research shows that the endowment effect weakens substantially when the owner is forced to evaluate the owned object against a clear, specific alternative rather than in isolation. The librarian who is asked "would you rather have this book that hasn't been borrowed in fifteen years, or the shelf space it occupies?" is using a different reference point than the librarian who is simply told "we should remove some books."

For a subscriber list, this means the conversation in the meeting room needs to change its frame.

Not: "should we remove these 200,000 dormant subscribers?"

But: "would you rather have 200,000 dormant subscribers who are damaging the inbox placement of your 480,000 active subscribers — or would you rather have 480,000 active subscribers reaching the primary inbox of people who actually want to hear from you?"

The same list. Two different reference points. Two very different endowment effect calculations.

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/7a39177c-5584-492a-9c6c-dd27c57a94c0.png align="center")

* * *

### The Practical Steps — Built Around the Reference Point Fix

**Step 1 — Segment before suppressing**

Do not look at a list of dormant subscribers as a single block to remove or keep. Segment by recency and type of last engagement:

```plaintext
Group A: No open in 12-18 months
         — Re-permission campaign first
         — One email, clear subject line: 
           "Do you still want to hear from us?"
         — Suppress anyone who does not engage 
           with the re-permission email itself

Group B: No open in 18-24 months
         — Suppress immediately from 
           all active campaigns
         — Hold in a separate suppressed 
           segment for 90 days
         — Then remove permanently

Group C: No open in 24+ months
         — Remove immediately
         — No re-permission attempt
         — The endowment effect on this segment 
           has already cost you more than 
           any conceivable future value
```

**Step 2 — Change the dashboard**

The metric that triggers the endowment effect is absolute list size — the number that goes down when you clean. Build a second view in your analytics that shows engaged list size separately from total list size, and track inbox placement rate as a primary metric alongside open rate.

When the dashboard shows "engaged subscribers: 480,000 / inbox placement: 87%" the cleaning decision looks like a win, not a loss.

**Step 3 — Set a dormancy threshold and enforce it automatically**

The endowment effect is strongest when the decision to remove is made manually, case by case, in a room where someone can argue for keeping each segment. Automate the suppression of subscribers who cross the dormancy threshold so the decision becomes a system rule rather than a recurring meeting agenda item.

The library that culls automatically on a ten-year-no-borrow rule never has to hold the meeting about whether to keep the books. The rule holds the line that the librarians' psychology could not.

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/26e354a3-34ac-4814-8db1-8054df295e95.png align="center")

* * *

### The Library That Finally Made Space

The library in this story did eventually commission a systematic cull.

It took eighteen months to complete. The decision was made not by changing the librarians' feelings about the books — those feelings did not change — but by changing the reference point. A new director presented the board with two futures: a library where every book was kept and the building's capacity to serve its actual borrowers continued to deteriorate, or a library where the space freed by the cull was used to create reading areas that attracted 40% more active borrowers within two years.

Same books. Same librarians. Different reference point.

The cull happened. 94,000 books were removed over eighteen months — donated to smaller libraries, digitised where possible, and in some cases simply retired. The borrowing rate of the remaining collection increased by 31% in the first year after completion.

The books that left were not valuable because the library owned them.

They were valuable to the borrowers who might have found them in another library, with more room to look.

Your list is the same calculation.

The subscribers who have not heard from you in two years are not valuable because they are yours.

*They are valuable to the brand that earns the right to reach them — when that brand is ready to send something worth opening.*

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/eaf83aab-3713-4436-b7cd-e49d836c4ec5.png align="center")

* * *

If your brain is already triaging this page for a 5-second window, skip the reading—the complete narrative is perfectly laid out in the infographic below.

![](https://cdn.hashnode.com/uploads/covers/6a22deaebdc172ef88447d5e/dcf131d4-6322-4400-96c2-45779cb92498.png align="center")

  

* * *

*Published by Hetvabhas — independent analysis of brand communication*  
*No vendor agenda. No sponsored content. No false reasoning.*