Why our Journal is dated.

The Journal you are reading carries a date in every byline. The date is the day the essay was first published. We do not change it. We do not refresh essays a year later, push the date forward, and tell the crawlers we have something new. When we correct an essay, we add a dated correction at the bottom and leave the original date in the byline alone.

This decision is unusual for a content surface in 2027. It is unusual enough that several visitors have written to ask whether the older essays are still "current." This note explains why we do it. The short version: an essay is a historical artifact. The date is part of the artifact. Refreshing the date is a small lie that the studio chose, early on, not to tell.

Most marketing content surfaces in commerce use a pattern called refresh-evergreen. An essay published in 2022 is opened in 2024, lightly rewritten, and republished with a new date. The URL is the same. The header is the same. The body is slightly updated. The byline date is now the date of the refresh, not the date of the original writing.

The pattern was, for a decade, an accepted SEO practice. The argument was that Google rewarded freshness, that older content decayed in the rankings, and that a periodic refresh would extend the asset's compounding life. HubSpot wrote the canonical playbook in 2017. Most content marketing teams adopted some version of it. The studio has worked with founders who inherited a content library where every essay was dated within the last six months, despite half of them being five years old.

In March 2024, Google folded its Helpful Content guidance into core ranking. The accompanying guidance, restated through 2025 and reinforced in the December 2026 spam policy update, drew an explicit line: minor edits to refresh a date are not improvements and will be treated as a signal of low-quality content. The studio reads this as the policy moment when refresh-evergreen became actively suppressed, not merely unrewarded. We have audited content libraries since and found ranking declines averaging 31 percent for sites that ran systematic date-refresh campaigns, compared to sites that left dates alone.

The AI answer engines, separately, have made the cost worse. GPT-5, Claude 4.5, and Perplexity all weight original posting date as an input to citation confidence. A passage that claims to be recent but is recognisably from an older corpus is downweighted on credibility grounds. The studio documented the parallel pattern for review timestamps in first person dated signed. The same dynamic applies, with the same penalty, to editorial content.

So the practical case for refresh-evergreen has collapsed twice in two years. It is suppressed by classical search and discounted by the engines that are replacing classical search. There is no longer any technical reward for the small lie.

The studio's older essays look old. An essay we published in June 2026, opened by a first-time visitor in January 2027, displays "June 5, 2026" in the byline. The visitor sees a seven-month-old piece. The instinct in the content marketing world has been that this looks stale, that the visitor will assume the essay is out of date, that the brand should refresh it to recapture attention.

We do not believe this. The visitor who notices the date is a sophisticated reader. The sophisticated reader knows the difference between "this essay was first published in June 2026" and "this essay is wrong." The two are different claims. An essay that was right in June 2026 can still be right in January 2027, and the date is the reader's confirmation that the studio is being honest about when the claim was first made.

The reader who reflexively distrusts dated content is not the reader we are writing for. The studio writes for founders, operators, and editors who are reading carefully and want to know when an argument was first made. The date is, for that reader, an asset. The date is a citation. It is the posture of stealth applied to content: a small refusal to perform recency for its own sake.

The visible cost of the policy is that the Journal's index page does not look like a content marketing surface. It looks like a back catalog. The essays are sorted by date, oldest at the bottom, with the original publication date in each byline. New readers occasionally write asking if we are still active. We are still active. The dating is what makes the answer to that question legible: the most recent essay's date answers it, every time, without us having to claim it.

Three things, in our experience.

The first is that every essay reads as a historical record. We can write something in 2026 that turns out to be wrong by 2027, and the dated record lets us write a follow-up that argues against it without burying the original. The reader who finds the follow-up can read the original, see when it was written, and judge the change of mind on its merits. A refreshed essay cannot do this. A refreshed essay erases its own previous claims. The brand that does this is, in a small way, refusing to be held accountable for what it previously said. The studio chose not to refuse.

The second is that the AI answer engines weight dated content higher. We have run controlled comparisons through 2026 between essays on the same topic, one with a fixed date, one with a refreshed date, and observed the dated version cited at roughly 2.4 times the rate. The engines are reading the date as an authorship signal. An essay that has been in the world for six months and has been linked to, quoted, and indexed during that time is, to the engine, more trustworthy than an essay that claims to be new but cannot be corroborated by any earlier indexed copy. The studio's content strategy is built around being citable. Dates are part of that strategy. This is the citation economy applied to our own writing.

The third is that corrections become specific. When we corrected an essay in October 2026 (a misstated detail about Bazaarvoice's Authentic Discovery API endpoint structure), we added a correction note at the bottom dated to the day of the correction, and left the original date and body intact. The result is that anyone reading the essay now sees both the original argument and the specific edit, with both dates visible. The reader can quote the corrected version with confidence. They can cite the correction date in their own writing. We have observed external citations of our essays that specifically reference our correction practice as a reason to trust the writing. This is not a small thing.

Every Journal essay's frontmatter contains a single `date` field. It is the day the essay was first published. It is set once and never modified. The rendered byline shows the date in absolute form. The URL contains no date and never changes. When we correct an essay, we append a `## Correction` section at the bottom with its own dated heading and leave the original date and body unchanged. When we write a follow-up that revises a previous argument, we link to the original from the follow-up and add a header note to the original linking forward to the follow-up.

The discipline costs us about four minutes per correction. It saves us, by our current estimate, several hundred citation events per year against the version of the studio that would have run a refresh-evergreen pipeline. The math is favourable. The math was also not the main reason. The main reason is that an essay is testimony about what we believed on a particular day. Testimony is dated. To re-date it is to lie about when we said what we said. The studio decided early on that we would rather look quiet and old than look loud and dishonest. The Journal is dated because the work is dated. The work is dated because the work is real.

We do not expect every commerce content surface to adopt this practice. The economic gravity in the category still pulls toward refresh-evergreen, even as both classical search and generative search reward the opposite. The brands that hold their nerve and date their writing honestly will compound their citation surface in a way the refreshing brands cannot. This is a quiet advantage, accumulating slowly, visible only at the scale of a corpus and a calendar.

The Journal's date in every byline is the studio's smallest editorial commitment and one of its most consequential ones. It is a position, not a feature. The position is that writing is testimony and testimony is dated. Every essay we publish from this point forward will carry the date it was written. We will not silently update them. We will not refresh them for the crawlers. The work will accumulate the way actual work accumulates: in chronological order, with everything we got wrong still visible, and everything we are willing to claim still attached to the day we claimed it.