How to Update Old Blog Posts for SEO: What to Change, What to Leave

Search for advice on how to update old blog posts and you will meet the same tip near the top of almost every result: republish with a new date. Several pages call it the easiest win available. One suggests adding the current year to a heading while you are there.

Google has answered that one in writing. Its guidance on helpful content lists, among the questions worth asking about your own work: "Are you adding a lot of new content or removing a lot of older content primarily because you believe it will help your search rankings overall by somehow making your site seem 'fresh?' (No, it won't)" [1].

The parenthesis is Google's, not mine.

So this guide is about the other thing, the part that does work and takes longer. To update old blog posts in a way that moves anything, you have to decide which post to refresh, find the actual gap, and change only what needs changing. The previous version of this page covered nine steps, seven of which are covered properly elsewhere on this site. Those are now links rather than summaries, and what remains is the part nothing else here covers.

Why old posts lose traffic

A page rarely gets worse on its own. What happens is that everything around it changes.

Competitors publish something more complete. The query itself shifts, so the kind of page Google wants for it is no longer the kind of page you wrote. Facts in the article go stale while the writing stays confident. Newer posts on your own site absorb the internal links the old one used to get.

None of that announces itself. Traffic does not fall off a cliff; it drifts down over months, and by the time the drop is obvious in a dashboard it has usually been running for two or three quarters. That is the real argument for a scheduled review rather than a reactive one.

Step 1: Pick the right post

This is the step that decides whether the work pays off, and it is the one most guides skip past.

Open the Performance report in Search Console [3], set the window to the last three months, and sort by impressions. What you are looking for is not your best page or your worst, but the ones in a specific band.

The band from 5 to 20 is where refreshes work because Google has already decided the page belongs in the conversation. Something specific is keeping it below the pages above it, and specific problems can be found.

Two other things worth sorting for while you are in there. Pages with high impressions and a low click rate have a snippet problem rather than a content problem. Pages that used to perform and no longer do are decaying, and the content audit guide covers how to spot that pattern across a whole library rather than one page at a time.

If Search Console is not set up yet, that is the prerequisite, and the Search Console setup guide covers it.

Step 2: Find the actual gap

Do not start editing. Start comparing.

Search your target query in a private window and open the top three results. Then work through four questions, in this order:

Is it the same kind of page? If the results are comparison posts and yours is a how-to, no amount of editing fixes that. This is the one failure that makes the rest of the exercise pointless, and it is worth settling before anything else. The search intent analysis guide goes into how to read a SERP for format rather than topic.

What do they cover that you do not? List every H2 and H3 they carry that your page lacks. Then cross off the ones that are padding. What survives is your gap, and it is usually two or three subtopics rather than two thousand words.

Is anything they offer missing from yours? Tables, worked examples, original data, screenshots, a calculator. Format is content. A page that answers the same question more usably is the better page.

What is stale in yours? Prices, product names, version numbers, screenshots of interfaces that have been redesigned, statistics with no date, links that now redirect or 404.

That list is your scope of work. Everything outside it stays as it is.

Step 3: Change what needs changing, and nothing else

The most common way a refresh goes wrong is enthusiasm. Someone opens an old post, decides the writing is dated, rewrites it end to end, and loses the sections that were doing the ranking.

Two principles sit behind that table. Age is not a defect, and a sentence written two years ago that is still accurate and still clear needs nothing. And every change you make is a variable, so the more you change at once, the less you will be able to say about what worked.

Step 4: Then fix the supporting pieces

Only once the content gap is closed. These are real, and they are second, because a well-linked page that does not answer the question still does not answer the question.

• The snippet. If impressions were high and clicks were not, the title and meta description are the problem rather than the body. The on-page SEO checklist covers what makes a title honest as well as clickable.
• Internal links in. A refreshed page deserves links from your stronger related pages. Internal linking strategy covers how to find which of your pages currently point at it, which is usually fewer than you expect.
• Internal links out. Point to whatever you have published since, which is often the real reason an old post feels dated.
• Schema. If the page qualifies for a rich result and does not have the markup, add it. Structured data for bloggers covers the types worth bothering with.
• Topic coverage around it. Sometimes the page is fine and the surrounding cluster is thin, which is a topical authority problem rather than a page problem.

Step 5: Set the date honestly, and record what you did

Change the modified date because the content materially changed. That is the whole rule.

Google's guidance on publication dates is that "The dates must describe the publication or update date of the page" [2]. A date moved forward on a page that gained a new paragraph and a swapped screenshot is not describing an update so much as performing one.

The year in a title follows the same logic. It earns its place when the content is genuinely annual, such as pricing or a version comparison. On an evergreen guide it creates a maintenance obligation, and a title reading 2024 in 2026 costs more trust than no year would have.

Then write down what you changed and when. Not for Google, for you. If the page moves in six weeks and you changed five things, you will want to know which five. If you are tracking this across a library, the content ROI framework covers what your analytics can and cannot actually attribute.

Step 6: Schedule the next one

A refresh is maintenance, not a project, and the difference shows up about eight months later.

A workable rhythm for most sites: update old blog posts in the top twenty by impressions once a quarter, and review everything else once a year. Put it in a calendar with a name attached. The habit is what catches decay early, because nothing else will.

For distributing the refreshed post once it is live, rather than leaving it to be rediscovered, the content distribution checklist covers the mechanics.

A note for Nepali bloggers

Local-language and local-market posts decay differently. The competing set is smaller and changes more slowly, which means an older post can hold its position far longer than it would in a crowded English market, and it also means a single well-executed refresh can be decisive.

The practical difference is in step 2. There are often fewer than three serious competing results to compare against, so the gap analysis turns into a question about what readers need rather than what rivals published. Keyword research for a Nepali blog and the blogging platform comparison for Nepal cover the groundwork.

Mistakes that cost traffic when you update old blog posts

• Rewriting what was working. The single most expensive error, and the easiest to make.
• Changing the URL. It throws away the page's history for a cosmetic gain. If the slug is genuinely wrong, that is a deliberate migration with redirects, not part of a refresh.
• Changing everything at once. You will never know what worked.
• Refreshing pages that never ranked. Those have an intent or eligibility problem that more words will not touch. Check the technical SEO basics before assuming content is the issue.
• Treating the date as the deliverable. Covered above, and worth repeating, because it is the most common advice on this topic and it is wrong.
• Checking the result the next morning. Recrawling takes as long as it takes, and a week of data proves nothing either way.

The bottom line

The page you already published is usually worth more than the one you have not written yet, because it has history and because the gap between it and the results above it is often small and findable.

Find the post in the 5-to-20 band. Compare it against what ranks. Close the specific gap. Change the date because the content changed, and not for any other reason.