Guides
Part of No shortcuts here: June birthdays turn on the basics
Why do famous June birthdays pages get copied instead of checked?
Famous June birthdays: the economics that make anniversary pages cheap to produce and not worth checking, and what a different set of choices looks like.
Anniversary pages exist in enormous numbers for a reason that has nothing to do with anniversaries. They are cheap to produce, they refresh themselves every year, and they attract searches that recur on a schedule. Understanding the economics explains most of what is wrong with them.
What to take away
- A page tied to a recurring date has predictable traffic, which makes it worth producing at volume and not worth checking.
- Volume production and verification pull against each other, and the economics decide the winner.
- The cost of being wrong falls on the person described rather than on the publisher, which is why the incentive does not correct itself.
What a June list actually holds
A June birthdays page is a fixed roster, and the same names turn up on nearly every version: Marilyn Monroe on June 1, Morgan Freeman on June 1, Angelina Jolie on June 4, Prince on June 7, Kanye West on June 8, Johnny Depp on June 9, Natalie Portman on June 9, Donald Trump on June 14, Paul McCartney on June 18, Lionel Messi on June 24 and Ariana Grande on June 26.
A copier does not need to know any of those names. The format does the work: a name, a day, a month heading. That is also why a wrong version spreads as easily as a right one. A page that misstates a day is copied as readily as a correct page, and once the same mistake sits on dozens of pages, repetition begins to read as agreement.
The dates themselves are not hard to check. Most of these lists trace back to a handful of large reference databases, Wikipedia and IMDb among them, where a reader can look up any one of the dates above. A copier never does.
The business model in one paragraph
A content farm produces large quantities of material aimed at search demand rather than readers, keeping production cost as low as the format allows.
Dated material is ideal input: demand is known in advance, repeats annually, and one template fills a year. The output need not be good, only exist and be found.
Publishing at volume is not new and is not by itself a problem. What changed is that the marginal cost of another page fell to nearly nothing while the marginal cost of checking one did not fall at all.
Why checking loses
Search engine optimization is the practice of shaping a page so that it is found, and nothing in it rewards accuracy directly. A page is rewarded for matching a query, for being usable, and for signals that suggest it is worth showing. A wrong date matches the query exactly as well as a right one.
Production choices and their effects
Production decision
- Copy details from another list
- Large saving
- Verify each detail against a source
- Large cost
- Add more entries per page
- Small cost
- Publish for every day of the year
- Multiplies output
- Remove an entry on request
- Small cost
Effect on cost
- Copy details from another list
- None
- Verify each detail against a source
- None directly
- Add more entries per page
- Positive
- Publish for every day of the year
- Positive
- Remove an entry on request
- Slightly negative
Effect on being found
- Copy details from another list
- Verify each detail against a source
- Add more entries per page
- Publish for every day of the year
- Remove an entry on request
Effect on cost
- Copy details from another list
- Large saving
- Verify each detail against a source
- Large cost
- Add more entries per page
- Small cost
- Publish for every day of the year
- Multiplies output
- Remove an entry on request
- Small cost
Effect on being found
- Copy details from another list
- None
- Verify each detail against a source
- None directly
- Add more entries per page
- Positive
- Publish for every day of the year
- Positive
- Remove an entry on request
- Slightly negative
Every row points the same way. This is not an accusation about anyone's motives. It is arithmetic, and it predicts the behavior you can observe.
What it does to the record
Copied details look like confirmation. Twenty pages with a value resemble twenty sources. A reader who checks three has checked one thing three times.
The January birthdays index sets documentation, not recognition, as the right threshold for a list, and copying machinery is exactly what that threshold keeps out.
The person described has no lever. They did not supply the detail, they cannot correct twenty publishers, and the one that responds is replaced by the copies. The February index covers why a birth date in particular is the field where this hurts, and the June index covers what a month grouping can honestly claim.
Making a different set of choices
- Publish fewer entries and say where each came from.
- Prefer a description that is checkable, such as a body of work, over a detail that is merely repeatable.
- Leave a gap visible rather than filling it from another list.
- Put the inclusion rule and the correction route in public, as the January birthdays index does, so that a reader can see what was applied and act on it.
- Accept that this produces a smaller page. That is the trade, and it is the only one that changes anything.
Common questions
Are all anniversary pages like this?
No. The format is fine and some are carefully made. The point is that the economics reward the other kind, so a reader should assume nothing from the format alone and look for stated sources.
How can a reader tell the difference quickly?
Look for whether the page says where anything came from, whether it admits to gaps, and whether it offers a route to a correction. Pages produced at volume rarely carry any of the three, because each one costs something and none of them helps the page get found.
Does search work against accuracy?
It is indifferent to it, which under volume production has the same effect. Accuracy is not a signal that can be measured cheaply, so it is not one that gets rewarded.
Is a smaller, sourced page actually useful?
To the reader who wanted to understand something, yes. To the reader who wanted a name and a number, no, and that is a reader this site cannot serve without doing the thing it exists to avoid.
