Every year there’s this scene: first week of peak season the boss asks “why isn’t our page ranking,” and you just start writing. Seasonal keyword competition ends before search volume even starts climbing — wait until the curve tilts up to act, and the new page hasn’t even stabilized indexing, let alone made the top 10.
Quick conclusion: moving the start date from “peak season month” to “90 days before peak season” is the watershed for whether you can eat seasonal traffic. In those 90 days, first 30 days build pages and internal links, middle 30 days use peripheral long-tail to push page authority, last 30 days do title and content fine-tuning. With three data sources — trend tools, historical traffic, competitor publish dates — you can lay out a 12-month search calendar in one go.
Why 90 days is the watershed, not 30
A new page from publish to stable ranking usually goes through three stages: indexing (3-14 days), initial ranking fluctuation (2-6 weeks), entering stable range (after 8 weeks). Add these three segments together and it’s close to 12 weeks, exactly 90 days. Pages started 30 days early are often still in violent ranking fluctuation during the two peak weeks — watching traffic flow right past you.
Search behavior itself is also moving earlier. Take categories with long decision cycles like renovation, study abroad, tax filing — the peak of research-type queries like “how to choose” “how much” is usually 6-10 weeks earlier than transaction-type queries like “where to buy” “nearby.” What you’re actually grabbing is the research-period entry point, not the last search on transaction day. The division of labor between these two types of keywords can reference search intent four-category judgment standards, landing research keywords and transaction keywords on different pages respectively.
Use three data sources to draw your annual search calendar
- Trend curves: throw 30-50 category keywords into trend tools, look at monthly indices over the past 5 years, record each keyword’s rise month, peak month, decline month. Months where at least 4 of 5 years overlap count as a pattern, peaks that only appeared once treated as noise.
- Your own historical traffic: export monthly clicks from the past 2 years from webmaster backend, break down by landing page. You’ll find some pages’ peak season isn’t synchronized with industry trends — these deviations often come from your own campaign rhythm, need separate annotation.
- Competitor publish dates: crawl first publish date and most recent update date of seasonal pages from 3 main competitors. How early competitors generally start is the actual competition starting line for this category, more accurate than any theory.
- Merge into one table: horizontal axis is 12 months, vertical axis is keyword groups, each cell filled with one of five states “start / build / promote / harvest / dormant” — at a glance you can see which months need extra manpower.
90-day rhythm table: what to do and what to watch at each stage
| Stage | Time from peak | Core action | Acceptance signal |
|---|---|---|---|
| Page building | Day 90 to 61 | Complete main page and 3-5 long-tail peripheral articles, lay internal links | All pages indexed, main keyword enters top 50 |
| Rank nurturing | Day 60 to 31 | Add cases, data, charts, add links from high-authority on-site pages | Main keyword enters top 20, long-tail starts getting sporadic clicks |
| Sprint | Day 30 to 8 | Change title and first screen per newly emerged search terms that year, update price and time info | CTR higher than site average, impressions clearly rising |
| Harvest | 7 days before and after peak | Only do conversion optimization, forbid major structure and URL changes | Conversion rate and inquiry volume, stop watching ranking fluctuations |
| Review | 15 days after peak | Record actual rise date and actual peak date, backfill next year’s calendar | Calendar error narrowed to within 5 days |
Forbidding major changes during harvest period is a pit many teams have stepped in. Changing URLs, switching templates, rearranging H2s mid-peak season equals asking search engines to re-evaluate this page when you most need stability — dropping 3-5 positions is enough to cost you half your earnings. Everything you want to change goes into the review list, wait for off-season to act.
Write seasonal pages as evergreen pages, not one-time materials
The biggest waste of seasonal pages is redoing them every year. This year write “how to choose for 2026 Double 11,” next year the whole article becomes invalid, authority resets to zero. The correct approach is one keyword group keeps only one long-term URL, no year in the path, year in the title placed in a replaceable position, current-year data in the body concentrated in 1-2 sections, only update those blocks each year. The correspondence between URLs and main keywords can be managed by table according to keyword to URL mapping method, avoiding new hires next year building another duplicate page.
Peripheral long-tail also needs evergreen-ification. The 3-5 supporting articles written around the main page should choose angles that don’t expire with the year like “how to judge” “what types are there” “common misconceptions” — the question-word combination formula mentioned in long-tail keyword research method can quickly put together a batch of topics that both have search volume and don’t expire. These articles continue eating traffic in off-season, and in peak season concentrate authority to the main page.
Off-season isn’t downtime, it’s time to stockpile ammo
Seasonal categories’ off-season is usually 5-7 months. Search volume is low this period, but competition is also loosest — exactly the window to铺满 peripheral content and straighten out internal link structure. Specifically do three things: fill in blank points in keyword groups that don’t yet have pages承接; add cases and charts to pages that performed well last year; clean up duplicate pages competing for the same keyword. The overall methodology can reference the quarterly cycle in keyword research playbook, treat off-season as the construction period for the next round of peak season.
If your business has regional attributes, off-season is also suitable for filling in regional keywords. Three-part combinations like “city + category + season” don’t have large search volume, but conversion rate is usually higher than generic keywords, and almost nobody has systematically done them —属于 low-cost high-return gap-filling actions.
Landing is four steps: open the calendar, find your 3 most important seasonal peak months, count back 90 days to mark start dates and set reminders; export 5-year trend data for 30 category keywords,圈出 rise months overlapping in 4+ years; check whether existing seasonal page URLs contain years, if so immediately plan into year-free long-term paths; for the nearest peak season, list the main page plus 3 peripheral articles, assign to people within this week. After one seasonal cycle runs through, backfill real data, next year’s error can be compressed to within a few days.
FAQ
How far in advance should seasonal keywords be laid out?
90 days in advance. A new page from publish to stable ranking takes 8-12 weeks, first 30 days build pages and internal links, middle nurture authority, last fine-tune title — wait until the curve tilts up and it’s too late.
How to judge whether a seasonal keyword truly has a pattern?
Pull monthly indices over the past 5 years, at least 4 years’ rise months overlapping counts as stable pattern; peaks that only appeared once treated as noise, don’t build a separate page for it.
Do seasonal pages need to be redone every year?
No. One keyword group keeps one year-free long-term URL, concentrate current-year data in 1-2 sections, only update those blocks each year — authority can accumulate year over year.
What should be done during off-season?
Fill keyword group blank pages, add cases and charts to pages that performed well last year, clean up duplicate pages competing for keywords. Off-season competition is loosest, exactly the window to stockpile content and internal links.
Can you change the title mid-peak season if it’s not working?
During harvest period forbid major changes to URLs, templates, and H2s — it makes search engines re-evaluate the page, rankings drop 3-5 positions. What you want to change goes into the review list, wait for off-season to act.
Figure: Key points of seasonal keyword trends: lay out search peak season 90 days in advance (compiled by 运营GO)
