A page can lose half its visits after a busy season and still be doing its job. It can also look healthy against a quiet month while performing badly at the point when readers need it most.
I compare seasonal content against the stage of demand it is serving. Calendar dates are one view; the weeks before an event, a deadline or a buying season are another. Keeping those views separate helps me avoid turning a normal seasonal drop into an unnecessary rewrite.
Describe the season before choosing the dates
I start by writing down what brings the audience to the page. Is it an annual event, a moving holiday, a recurring deadline, a weather-dependent task or a product people buy at a particular time?
I also identify the point at which the page is useful. A preparation checklist might attract readers weeks before an event, while a results page becomes useful afterwards. Combining them in one topic total can hide that they serve different parts of the journey.
My note includes the event or season dates in each year, the intended audience and anything unusual: a cancellation, a changed format, a different market or a temporary surge in publicity. If those changed, I keep them beside the figures rather than pretending the two periods are interchangeable.
Keep three comparisons available
I usually want three views, each answering a different question.
- The preceding comparable period: what changed recently? Equal complete weeks help limit differences caused by weekdays.
- The same calendar period last year: is there an annual pattern? This is useful when the season falls at roughly the same time.
- The same stage before or after the event: how did the content perform when it was serving the same immediate need?
The third is my analytical method, not a special Google score. I choose an anchor such as the opening date or submission deadline, then write down the exact interval: days 28 to 1 before that anchor, for example. I apply the same rule to both years.
If that produces a different weekday mix, I record it. Sometimes complete matched weeks are more useful; sometimes the exact event stage matters more. I choose deliberately and retain the other comparison as context.
Use complete data and unchanged filters
The comparison needs the same property, search type, audience filters and page scope. I do not compare a country-specific mobile report with last year’s all-country, all-device total. When a URL has changed, I record which old and new pages belong to the same content journey.
I inspect the date boundaries as well as the chart label. Google’s Performance report grouping guidance notes that first and last weekly or monthly groups can be partial. A line labelled with a month is not necessarily a complete month’s measurement.
I keep collection dates and unavailable periods visible. If last year’s data was not retained, I cannot reconstruct it by copying a current number into an older column. A shorter honest comparison is more useful than a fabricated annual trend.
A fictional example: the event moved
These dates and figures are invented. Imagine an event held on 15 June one year and 29 June the next. An article helps visitors prepare during the four weeks beforehand.
A fixed May comparison could make the later event look weak simply because fewer people have reached that preparation stage. For an event-relative view, I compare 18 May–14 June in the first year with 1–28 June in the second: two complete 28-day windows immediately before the respective opening dates.
Suppose those windows produced 1,000 and 920 search clicks. The second is 8% lower. That is the observation I would investigate; I would not use a much larger fall from an unrelated post-event period as the headline.
I would still check impressions, useful queries, device and country mix, and the article itself. The later event may have different demand, and an apparently similar event may attract a different audience. Matching the dates removes one obvious distortion; it does not make the comparison a controlled experiment.
Check whether demand changed beyond your site
If the relevant queries are quieter, I look for wider context. Google Trends can help show whether interest in a topic has shifted, but it does not tell me the exact number of searches my page should receive.
Google’s explanation of Trends data describes sampled, normalised relative interest. A value of 100 is the peak within the selected comparison, not 100 searches. Low-volume terms can also show zero, so I do not read that as proof nobody cares about the subject.
I keep the geography, term or topic choice and time range consistent. Then I ask a narrow question: does this support a broader demand change, or does the loss seem more specific to my page? I still need the page and query evidence before choosing an action.
Give seasonal pages a useful quiet-period job
A quiet month can be the right time to check dates, links, accessibility, booking instructions and anything that became misleading after the last event. I want the next reader to understand which information is current and which describes a previous edition.
I do not automatically replace a useful historical page with next year’s promotion. Sometimes a stable recurring guide should be updated; sometimes an event-specific record needs to remain clearly dated alongside a separate current guide. The right choice depends on the questions people are trying to answer.
Before combining pages or redirecting an old edition, I check its remaining searches and links, then verify the replacement answers that need. The content-decay review has a decision table for keeping, updating, combining, archiving or withdrawing content.
Record the comparison so somebody else can repeat it
My seasonal review note is short enough to read beside the spreadsheet:
- The page or page group, its purpose and the intended audience.
- The event or season anchor in each year, and why those anchors are comparable.
- The exact inclusive start and end dates, time zone and number of complete days.
- The property, search type, country, device and any query or page filters.
- The collection date, missing data and material changes in the event or site.
- The absolute and percentage differences, followed by the next check.
I save the calendar and event-relative views rather than replacing one with whichever tells the better story. When I make an improvement, I also set the next useful review point; checking a content refresh is easier when that decision is made before the result arrives.
If seasonal reports keep producing contradictory advice, I can help build a clearer SEO review and measurement process. The aim is to make the next decision fit the audience’s needs at that point in the season.