NovAsia

A fresh page can still contain old data

Ekaterina Andreeva on separating the period a number describes, the source publication date and NovAsia’s own review date.

This article reflects the named expert’s practical perspective. See NovAsia’s editorial policy for how material is prepared and reviewed.

A page can be updated today and still contain a number that describes last year. That is not automatically a problem; hiding the distinction is.

On NovAsia’s market-data pages, a useful number needs more than a fresh page timestamp. I want to know the period it measures, when the source released it and when our team last checked the claim. Those dates answer different questions.

An August report may summarise transactions from January to June. A current project card may use a price checked this week while the underlying campaign began earlier. If we flatten those dates into one “updated” label, the reader loses the chronology.

The temptation is understandable. Interfaces like one clean date. Readers like to know that a page has been reviewed recently. But a single timestamp cannot describe every fact inside an article. A legal rule may remain unchanged for years, an asking price can move quickly, a construction milestone changes on a different schedule again, and an annual market report will always look “old” compared with live inventory even when it is the newest official dataset available.

I find it more useful to think in layers. There is the date or period of the underlying event, the date the source published it, and the date the editorial team last verified that the claim still belongs in the article. Sometimes there is a fourth layer: the date a source page itself was modified, which may or may not mean the particular number we use was changed.

Fresh publication, older observation

This matters most when two facts are placed next to each other. A newly released macro report and a live asking price can look contemporaneous even when their measurement periods are months apart.

Consider a comparison table with an average market price, the current asking price of a particular unit and the latest transaction volume. All three may have been collected on the same day. They still describe different moments. The market average may cover a quarter, the asking price is a snapshot, and transaction volume may be published with a reporting lag. Without dates or periods beside them, the table creates a false impression of simultaneity.

On NovAsia, the research and market material separates the observation period from publication and review for this reason. The point is not to make every old source disappear. Sometimes the older source is still the best available evidence.

Wording matters as much as metadata. “The price is $X” claims something about the present. “The developer’s price list dated March showed $X” makes a historical claim. The source can be identical in both cases, but the editorial burden is different. If we no longer have recent confirmation, changing the sentence may be more accurate than pretending a new page update refreshed the old number.

Legal and regulatory sources create the opposite problem. A rule published years ago may still be current, while a blog post written yesterday may already be wrong if it relied on an outdated version. Here, age alone is a poor proxy for quality. The real question is whether the instrument has been amended, replaced or superseded by later guidance.

Source pages can also be misleadingly fresh. A website may show a new “last updated” date after a layout change, while the price table or project status remains untouched. That is why I try to verify the date of the claim, not simply the date printed somewhere on the page.

For market commentary, publication lag deserves its own explanation. Official numbers are often valuable precisely because they are measured carefully, and careful datasets can arrive later than marketing or listing data. That does not make them useless. It means they should answer historical or structural questions, while current inventory answers a different one.

I also avoid using a recent article date to lend freshness to an old observation. Republish or revise a piece as often as necessary, but the age of the evidence should remain visible where it affects interpretation. A reader should never have to infer that a 2026 page is relying on a 2024 market snapshot for a statement written in the present tense.

In that case I would rather label its age clearly than let a new article date make the underlying fact look newer than it is.

Good date handling is not about displaying as many timestamps as possible. It is about letting the reader reconstruct the timeline with minimal effort. If they can tell when something happened, when it was reported and how recently the claim was checked, they can judge whether the fact belongs in a decision today or only in the background story.