Keep the original publication date and add a separate last-reviewed or last-updated line saying what changed. Two dates tell the truth where one cannot.

The temptation

An old page still ranks and still gets read, and its date is three years ago.

The instinct is to change the date, which takes a second and makes the page look current.

Sometimes a word is changed first, so that something technically happened.

The page now claims to be from this month and says exactly what it said in 2019, which is a claim about currency that is not true.

Why it matters

The first is the practical one. A page dated last month that refers to something as recent when it was four years ago tells a reader the date is decorative, and they discount every date on the site afterwards.

Two dates solve it

The arrangement that is both honest and useful.

Keep the original publication date, which is a fact and does not change.

Add a second line: last reviewed, or last updated, with the date of the actual review.

Where something substantive changed, say what in a few words: prices updated, revised after a regulation change, figures refreshed.

That gives a reader everything they need, which is when it was written, when somebody last checked it, and whether anything moved.

It also lets you review a page without changing it, and record that you did, which is a genuinely useful thing to be able to say.

Reviewed is a real claim

Worth being clear about, since a reviewed date can become as hollow as a changed one.

A review means somebody read the page and confirmed it is still accurate, not that a script touched the file.

Checking that the figures still hold, the process is still what you do, and nothing referenced has changed takes about ten minutes for most pages.

Do that, then set the date, in that order.

A reviewed date on a page nobody read is the same dishonesty in a more sophisticated form, and it is easier to believe you are not doing it.

A worked example

A business had thirty advisory pages, all showing dates from between 2017 and 2019, and had been advised to refresh them all.

Rather than changing thirty dates, they worked through six pages properly over a fortnight: checked figures, updated prices, removed two references to things that no longer existed, and added a paragraph to three of them.

Each got a last-reviewed line with the month, and the three with substantive changes said what had changed.

The other twenty-four were left with their original dates untouched.

Over the following months they worked through the rest at about four a month, which meant that within a year every page had a genuine review date.

The pages that had been genuinely improved performed better, and the ones that had only been date-stamped in the alternative plan would have done nothing.

When the date should change

Since this is not an argument for never republishing.

Where a page has been substantially rewritten, with a new structure and largely new content, changing the publication date is defensible because it is effectively a new page.

Where the topic has changed materially, such as a rule change that makes the old advice wrong, republishing signals correctly that this is now different.

The test is whether somebody who read the old version would find the new one materially different.

Where they would not, it is a review rather than a republication, and the two dates handle it properly.

Some pages should have no date

A distinction worth drawing, since not everything benefits from one.

Service pages, contact details, and descriptions of what you do are not time-bound and a date on them creates a staleness signal for no reason.

Guidance that does not change, such as how something works or what to check, also ages badly with a publication date and well with a reviewed one.

News, announcements, offers, and anything referring to a specific year need a date and are misleading without one.

Decide which category each page is in rather than applying dates uniformly because the system does it automatically.

Most site tools date everything by default, and turning it off for the pages that do not need it is a settings change.

Keep a record separately

Because the date on the page is for readers and you need something for yourself.

A simple list of pages with the date each was last genuinely reviewed lets you work through them systematically rather than by whichever you happen to remember.

Note what was checked, briefly, so a review next year knows what the previous one covered.

That also tells you which pages have never been reviewed, which is usually more than anybody expects.

Four pages a month clears thirty in under a year and is small enough to actually happen.

The counter-case

Dates matter less than this treatment suggests.

Most readers do not look, most of the time, and a good page with an old date outperforms a poor page with a recent one comfortably.

There is also a real argument that publication dates on evergreen guidance do more harm than good and are better removed entirely than managed carefully.

And a business agonising over date policy while its prices are three years out of date has the effort in the wrong place.

Keep the original date, add a reviewed date when you have genuinely reviewed something, and spend the time on the review rather than on the date.

What to do

  1. Keep the original publication date.
  2. Add a last-reviewed line when you review.
  3. Say what changed if anything did.
  4. Actually read the page before dating it.
  5. Remove dates from pages that are not time-bound.
  6. Republish only after a substantial rewrite.
  7. Keep a review list and do four a month.

Step four is the whole of the honesty here, since a reviewed date applied without a review is the same claim in a form that is easier to tell yourself is different.

What staleness signals to a visitor is covered in a site last updated in 2017.


Frequently asked questions

What is wrong with changing the date?

The page then claims to be current while saying what it said years ago. Readers who check find the content gives it away, and they discount every date on the site afterwards.

What is the honest arrangement?

Keep the original publication date, which is a fact, and add a separate last-reviewed line with the date of the actual review and a few words on what changed.

What counts as a review?

Somebody reading the page and confirming the figures still hold, the process is still what you do, and nothing referenced has changed. About ten minutes for most pages.

When should the publication date change?

After a substantial rewrite with new structure and largely new content, or when the topic has changed materially. The test is whether a previous reader would find it different.

Should every page have a date?

No. Service pages and descriptions of what you do are not time-bound and a date creates a staleness signal for no reason. News and offers need one.

How do I keep on top of it?

Keep a separate list of pages with the date each was genuinely reviewed and work through four a month, which clears thirty in under a year.

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.

Thirty pages with old dates?

Review four properly this month rather than changing thirty dates. One of those produces something and the other produces nothing.

Start a Conversation