Old articles aren’t done when written — they decay over time. A content refresh schedule schedules checkups for each article by decay speed: fast-decaying quarterly, slow-decaying annually. Proactively keep them alive, instead of waiting until traffic is all gone before remembering to rescue. By then, rescuing often means rewriting most of it, and may not even get back the original rankings. Many sites put all their energy into new articles and let old articles slide — but old articles have a large base, and the loss from overall decay is often underestimated. Putting refresh into the calendar is using the least effort to preserve existing assets.
Set frequency by decay speed
First use historical decay curves to divide content into three tiers: fast-decaying time-sensitive content quarterly, medium-decaying semi-annually, slow-decaying evergreen content annually. Frequency is set by data, not by gut feel. Each article has its own rhythm, instead of one-size-fits-all same frequency for all. Accurate tiering means checkups won’t be either too frequent or too lazy — resources go where they count, no waste and no omission.
Decay must be actually measured, don’t rely on impression. Run through old articles’ traffic decline curves, and tiering has evidence. Fill checkup days in the schedule by tier, check when the time comes, don’t rely on remembering. Only then won’t old articles be forgotten in a corner — by the time you remember, it’s already lost a third of its traffic, and rescuing then is expensive. Measure once, and later scheduling is all easy, the archive can also be directly reused, and it gets more and more cost-saving with use.
After tiering, each article’s tier should be written into the content archive, set at publishing time. Publish = set tier, and later scheduling doesn’t have to guess how often this article should be checked. With the archive in hand, checkup days auto-pop reminders, and refresh truly gets done on time, instead of panicking to补救 after traffic drops. Set tier one step early, and the whole year’s refresh schedule goes smoothly — no need to临时 cram. Early tiering means the whole year’s refresh has evidence to rely on, no need for临时 judgment every time.
Checkups have a checklist
Each checkup goes through four points: is data dropping, is information outdated, are external links still alive, are internal links still working. Checklist ticking is more stable than reading the whole article, and doesn’t miss items. Whoever does it uses the same ruler, quality doesn’t fluctuate with the executor’s mood, and won’t skip an item because of rushing. The checklist turns checkups into replicable actions, not relying on individual发挥.
Checkups should be lightweight. Most articles done in half an hour: add a paragraph of new data, swap an example, fix a dead link. The focus is doing it regularly, not major revisions every time — major revisions反而 easily introduce new problems. Lightweight and high-frequency is more sustainable than heavy and low-frequency, and easier to persist into a habit. Treat it as a fixed action in the schedule rather than临时 firefighting, and old articles can keep living.
The checklist itself should also evolve with the site. When new content formats appear, add new check items; when old problems decrease, delete them. The checklist lives in the process, and checkups stay useful, won’t become an old table nobody trusts. Review the checklist once every six months,正好 aligning with the checkup rhythm. When the checklist is alive, checkups don’t go through the motions, and old articles’ lifespan is truly managed. Checklist always fresh, checkups don’t go through the motions, old articles truly get持续 care.
Refresh = re-publish signal
After major revisions, resubmit for indexing and do a round of internal linking, so the update is quickly sensed by search engines. Refresh isn’t self-amusement — you need to actively tell it “I’ve updated,” and authority flows back faster. Otherwise you quietly changed it, and it may not know to re-evaluate — the old article’s ranking may stay停 on the old version, the earlier effort equals wasted, and traffic doesn’t come back.
Re-publish signals should be written into the schedule. On the day refresh is complete, schedule internal linking actions, and authority flows back faster. Old articles therefore stay fresh, decay is压 down, and multi-year old articles can also steadily produce output. Asset lifespan is actively extended, not left to wither. Many sites ignore this step — refresh is done but doesn’t show effect, the root is missing this signal, and the update equals telling nobody.
Re-publish signals aren’t limited to internal links — it can also be an update note at the top of the article, telling readers what this version changed. The note makes readers feel the content is seriously maintained, and also helps search engines identify update points. Small action, low cost, but turns refresh from silent to有声 — the effect difference is big.有声 refresh is easier for both readers and algorithms to catch, and old articles are truly awakened. The gap between有声 refresh and silent refresh will be obvious after half a year.
Build a decay archive
Each article’s decay tier and last checkup day are recorded in the archive, and the schedule auto-reminds the next checkup. The archive turns refresh from “do it when remembered” into “must do when due” — old article management is sustainable, and won’t be lost when old-timers leave. The more complete the archive, the more worry-free the refresh schedule, the healthier the site overall, and experience沉淀 in the system rather than in someone’s brain.
The archive should link with new articles’ publishing dates. As soon as an article is published, set its tier and next checkup day — don’t drag until it’s old to补. The more complete the decay archive, the more worry-free the refresh schedule, the healthier the site overall, and old articles are no longer a forgotten burden but a foundation for持续 output. It’s one set with the publishing schedule, not two sets — at the moment of publishing, the checkup should already be scheduled, don’t wait for problems to补救.
The archive can be very simple — one table recording article name, publish date, decay tier, last checkup day, next checkup day. Simple means someone maintains it; complex反而 gets abandoned. The thicker the table gets with use, the more worry-free the refresh schedule, and old article management shifts from memory-based to system-based — not lost even when people leave. Simplicity is the key to the archive surviving — don’t build complex systems from the start. The table is simple, but it’s the底气 for old articles’ long-term health.
| Tier | Frequency | Action | Time |
|---|---|---|---|
| Fast decay | Quarterly | Add data | 30 min |
| Medium decay | Semi-annual | Swap examples | 1 hour |
| Slow decay | Annual | Major revision | Half day |
Figure: Four-step scheduling flow — Plan, Schedule, Execute, Retro
Key to making the schedule stick
Putting all of this into the calendar — the hardest part isn’t designing it, it’s sticking to it. I recommend first running a small rhythm smoothly, like first scheduling a checkup for the oldest batch of articles, then gradually adding items. To judge how often to check up, first look at the content decay curve; for more data-side decline measurement, you can对照 and set tiers in the data-side decay analysis.
