The Content Refresh Checklist: What to Change, What to Leave

The short version: a content refresh isn't a fixed list of thirty things you do to every old post. It's this: figure out why the page lost traffic, change only the thing that matches that reason, leave the rest of the page alone, and then check whether the clicks came back. Most "refresh checklists" skip the diagnosis and have you rewriting titles that were fine and padding pages that didn't need it.

Here's the checklist I actually run on my own posts — and, just as important, the part where you decide what not to touch.

Copy this checklist

Paste this somewhere and work down it per page:

CONTENT REFRESH CHECKLIST (per page)

Before you touch anything
[ ] Record the page's current clicks (last 28 days) and today's date
[ ] Note which query it's losing ground on

Diagnose why it slipped (pick the one that fits)
[ ] Position held, CTR fell   → it's a title/meta problem
[ ] Position dropped          → it's a ranking problem
[ ] Facts/dates/screenshots stale → it's a freshness problem
[ ] Thinner than what now ranks   → it's a depth/intent problem
[ ] Lost internal links / orphaned → it's a linking problem

Change only what matches the diagnosis
[ ] Title/meta problem  → rewrite the title + meta description, leave the body
[ ] Ranking problem     → add depth, match intent, refresh facts, add internal links
[ ] Freshness problem   → update stats, dates, examples, screenshots, pricing
[ ] Depth/intent problem→ add the sections competitors cover; consider a rewrite
[ ] Linking problem     → link to it from 3–5 strong, relevant pages

Leave alone
[ ] The URL (keep it)
[ ] Sections still pulling clicks
[ ] A title that's still winning its click
[ ] The publish date, unless the update is substantial
[ ] Pages that aren't actually decaying

After you publish
[ ] Wait 4–8 weeks, then compare clicks to the baseline you recorded

The rest of this is why each line is there.

Before you change anything: record the baseline

This is the step almost everyone skips, and it's the one that decides whether you ever know the refresh worked.

Before you edit a single word, write down two things: the page's clicks over the last 28 days, and today's date. That date is your edit marker — the "before" line that the "after" gets measured against. Without it, six weeks from now you'll be squinting at a graph trying to remember when you made the change, and you'll end up telling yourself it worked because you want it to have.

It takes about a minute. Open Search Console, filter to the one URL, note the number, note the date. Now you can prove the outcome later instead of guessing at it. (That measurement is a whole method of its own — I'll link it at the end.)

Diagnose why it slipped — then change only that

A page doesn't just "get old." It loses traffic for a specific reason, and the reason tells you what to change. Change the wrong thing and you can spend an afternoon rewriting a page that only needed a new title — or, worse, gut a page that was mostly working.

So before the checklist, do the diagnosis. In Search Console, compare the recent window to the page's healthy baseline and look at which number moved:

• Position held, but clicks and CTR fell. People see you in the results and scroll past. That's a title-and-meta problem, not a content problem — the cheapest refresh there is. Rewrite the title and meta description to be clearer and more specific, and leave the body alone. Rewriting 1,500 words because the title was dull is the most common wasted refresh I see.
• Position dropped down the page. You've genuinely lost ranking to someone. This is the real refresh: the content needs to be better than whatever passed you. Look at what now outranks you and close the gap — more depth, better match to what the searcher actually wants, current facts, a few internal links pointing in.
• Facts, dates, or screenshots are stale. The ranking is fine but the page references a 2023 number, an old price, or a screenshot of an interface that's changed. Update the specifics. This is low-effort and high-trust — an out-of-date stat is the fastest way to lose a reader who was about to believe you.
• The page is thinner than what ranks now. Search intent shifted, or competitors simply cover more. Add the sections they have and you don't. If the whole angle is now wrong, that's a rewrite, not a refresh (more on that below).
• It lost internal links or went orphaned. Pages you used to link from got deleted or restructured, and this one quietly lost its internal support. Re-link to it from three to five strong, relevant pages on your site. That alone can nudge a page that slipped for no on-page reason.

One diagnosis usually dominates. Fix that one first, ship it, and measure — don't do all five at once, or you'll never know which change actually moved the needle.

The changes themselves

Once you know the cause, the actual edits are the boring, well-established stuff — do the ones your diagnosis called for, skip the rest:

• Title and meta description: clearer, more specific, benefit-first. This is what most affects whether people click. Google's guidance on title links and snippets is worth reading once — it also explains why Google sometimes rewrites your title anyway.
• Facts and figures: replace anything more than a year or two old with current, sourced numbers. Nothing kills credibility faster than a page that calls itself current while quoting stale data.
• New sections: add the subtopics and questions the ranking pages answer and you don't.
• Internal links: add links to this page from relevant, high-traffic pages, and add links from it to your related posts.
• Images and screenshots: replace anything visibly dated; keep alt text accurate.

What to leave alone

Here's the half no checklist tells you. A refresh is surgical. The pages that lose ground from a refresh usually lose it because someone changed things that were working. Leave these:

• The URL. Keep it, almost always. The URL carries the page's age, its backlinks, and its ranking history — change it and you throw all three away and start the page over from zero. Improve the content on the existing address. Only change a URL when the topic has changed so completely it's genuinely a different page, and even then, redirect the old one.
• Sections that are still pulling clicks. If part of the page ranks for its own queries, don't "tidy" it out of existence. Check what the page ranks for before you delete anything.
• A title that's still winning its click. If CTR is healthy and position is holding, the title is doing its job. Leave it and spend the effort where CTR actually dropped.
• The publish date — unless the update is substantial. Bumping the date signals freshness when there's a real update behind it. Google rewards the substance of the change, not the date stamp, and it's built to notice a date that moved with nothing behind it — its helpful content guidance is explicit that changing dates to make content seem fresh is not the point. Never unpublish and republish to reset the date, either — that can wipe the ranking history you've spent months building. Update "last modified" when you genuinely changed the page; skip it when you didn't.
• Pages that aren't actually decaying. The biggest "leave alone" of all. A page holding its traffic doesn't need a refresh, and an evergreen post that's still ranking is a page you can ignore for another quarter. Refresh the ones losing clicks; leave the healthy ones to keep working.

Skipping well is half the method. Every page you don't refresh is an hour you get to spend on one that's actually bleeding.

Refresh, rewrite, or redirect?

Three different moves, one quick rule:

• Refresh when the page's structure is still right and the details went stale. Same URL, improve in place. This is most cases.
• Rewrite when the angle itself no longer matches what people search for — the page answers a question nobody's asking anymore. Rebuild from scratch, but keep the same URL so you don't lose its history.
• Redirect when the page is genuinely obsolete or overlaps another page so heavily they compete. Merge the useful parts into the stronger page and 301-redirect this one to it.

If you're unsure between refresh and rewrite, start with the refresh — it's cheaper, and the measurement afterward tells you whether it was enough.

There's a fourth move worth naming, and it's the one people skip: doing nothing. Some decaying pages shouldn't get any of the three above, because no version of them wins. When not to refresh a post walks that decision — and what to do with the hour instead.

Whichever you pick, once the page is live you can ask Google to recrawl it with URL Inspection in Search Console. It doesn't guarantee anything or speed up ranking — it just shortens the wait before Google sees the version you actually want measured.

After you refresh: check that it worked

You recorded the baseline before you started, so this part is easy. Wait four to eight weeks — rankings move in the first couple of weeks, but clicks lag, so measuring too early tells you nothing — then compare the same-length window after the change against the window before. If the clicks came back, you've got a repeatable playbook. If they didn't, you learned that for the cost of one afternoon instead of ten.

That verification step is a method in itself, seasonality traps and all — here's how to tell if a content refresh actually worked, start to finish. And knowing when a page genuinely isn't worth refreshing at all is its own decision worth making before you start.

Do the diagnosis part automatically

The checklist above works by hand — I'd rather you understood the diagnosis than pasted a title into a tool and hoped. But the slow part is the front half: finding the pages that are actually decaying, ranking them by the clicks at stake, and working out why each one slipped before you decide what to change.

That's the job RefreshRadar does. It watches every page in your Search Console, hands you the ranked list of the ones losing traffic, and tags each with the reason in plain English — position slip, CTR drop, staleness, lost query — so you walk into the refresh already knowing which line of this checklist applies. If you're weighing it against the other Search Console tools, the SEOTesting comparison covers what each one actually measures.

Connect Search Console read-only and see what's decaying on your site — free — no card, about 30 seconds, and we never post or change anything. If nothing on your site is decaying, we'll tell you that instead.