Pagination Done Badly

What happens beyond page one

A blog listing, a product category, or an archive splits into pages, and almost everything past the first is rarely visited by anybody.

Visitors do not click through to page four. Crawlers reach the later pages less often and treat them as less important, which is a reasonable judgement.

Which is fine for the listing itself, since nobody needs page four of a listing to rank.

It matters because for many sites the paginated list is the only route to the older items, and an item reachable only from page four is effectively hidden.

The three common faults

• Every page sharing one title and description.
• Pages loaded by a script with no real link.
• Later pages canonicalised to page one.
• No other route to the items on them.
• Addresses that change as new items are added.

The third is worth understanding: declaring page two as a duplicate of page one tells search engines to disregard page two, which also discards the links it contains to the items on it.

Real links, not a button

The most consequential of the faults, because it makes the later content genuinely unreachable.

A load more button or an infinite scroll that fetches items with a script produces no address and no link for a crawler to follow.

So the items beyond the first batch exist only for a visitor who scrolls, and are invisible to anything that does not run the script and scroll.

That is a real loss on a site whose older content is worth finding.

The remedy is not to abandon the button. It is to have both: a script for convenience, and real numbered links to real addresses underneath, which cost nothing and make everything reachable.

Distinct titles

A small fix that resolves a common reporting complaint.

Where every page of a listing shares a single title, reporting tools flag duplicate titles across a dozen pages, and search engines have nothing to distinguish them.

Appending the page number is enough: the listing name followed by page two.

Most platforms do this and some do not, and it is one line of configuration in a theme or plugin.

The same applies to the description, though titles matter more.

Do not attempt to write unique descriptive text for page seven of an archive, which is effort spent on a page nobody will read.

A worked example

A business had published around a hundred and forty articles over several years, listed twelve per page.

Their listing used an infinite scroll with no numbered links, and a plugin was declaring every listing page as a duplicate of the first.

Search Console showed the older articles being crawled rarely, and several had never been indexed at all.

The articles were good and were reachable only by scrolling a listing that produced no addresses.

They added numbered links beneath the scroll, corrected the canonical behaviour so each page declared itself, and added a simple archive page listing every article by year.

The archive page was the change that mattered: within two months most of the previously unindexed articles had been crawled and indexed.

Give items a second route

The structural answer, and it is more reliable than getting pagination perfect.

Anything that matters should be reachable from somewhere other than a paginated list.

Category pages, a plain archive listing everything, related links at the foot of each article, and links from your main pages to your best content all serve this.

A simple page listing every article, grouped by year or by topic, is unglamorous and effective, and takes minutes to produce on most platforms.

It also helps visitors, who are considerably more likely to browse a single organised list than to page through twelve at a time.

With that in place, pagination faults become cosmetic rather than costly.

Watch for shifting addresses

A subtler problem on listings that grow.

When items are added, everything shifts: what was on page two moves to page three, and the contents of every numbered address change.

Which means a crawler that recorded page three last month finds different items this month, and nothing about the address is stable.

That is unavoidable with chronological listings and it is another argument for the second route, since the item's own address never changes even as its position does.

It is also a reason not to link to a numbered listing page from anywhere permanent, since what it contains will not be what you intended.

How many per page

A choice usually left at whatever the platform defaulted to, and worth one minute of thought.

Too few per page produces many pages and buries older items deeper. Too many produces a heavy page that loads slowly, particularly where each item carries an image.

Somewhere around twenty is a reasonable middle for a text listing, and fewer where each item is image-heavy.

The practical test is the page weight: load your listing on a phone and see how long it takes, since a listing showing fifty products with fifty photographs is a performance problem wearing a navigation costume.

Where you increase the count, make sure images on the listing are the small versions rather than full-size ones scaled down in the browser.

The counter-case

Most small sites do not have this problem at all.

A business with fifteen articles and one page of listing has nothing to fix, and a shop with forty products across two pages has very little.

Pagination becomes worth attention somewhere above roughly fifty items, and below that the effort is better spent on the items themselves.

There is also a limit to how much a listing page deserves: these are navigational pages rather than destinations, and nobody should be optimising page six of an archive.

Add the second route, make sure real links exist, and stop there.

The check

1. Count your items and see how many pages that makes.
2. Check whether numbered links exist in the page source.
3. Confirm each page declares itself as canonical.
4. Append the page number to titles.
5. Build a plain archive page listing everything.
6. Add related links at the foot of each item.
7. Check whether older items are actually indexed.

Step five is worth more than the other six together, since it makes every item reachable regardless of how the pagination behaves.

Whether a listing is worth having at all is covered in a blog nobody asked for.

Frequently asked questions

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.