We Stopped Optimizing for Ten Months. Here Is the Content Decay Chart.

Anyone helping you with your SEO will show you a before-and-after chart. Here I'm going to do the same, but with an addition. I'm going to show you a before, an after, and an after-that — because the third act is the one nobody else measures, and it is the one that changes how you should think about optimization entirely. The industry has a name for what happens in that third act: content decay. Almost nobody shows you theirs.

The three acts

ScandinavianShoppe.com is a store I know as well as any store on the internet — I helped my best friends start it and still help them to this day. Based upon past experience, we built their business on years of meticulous manual SEO which rewarded the site with 500+ top-ten keyword rankings (and 1,500+ total ranking keywords), steady, for years.

Unfortunately, the nature of their business is that they only refresh a portion of their catalog every year with new products, and a few years ago Google started to see this as "diminishing content." The consequence of this was not completely unexpected:

• A reduction in ranked keywords

• A corresponding reduction in organic traffic (and total traffic)

• A corresponding reduction in sales

With the advent of AI, two years ago we began the process of updating their product catalog — product descriptions, page titles, meta data, the works. In one AI-assisted batch, we updated EVERYTHING...every single one of their 7,000 products. The following is what happened next.

Act one: the lift. An AI-assisted rewrite of the catalog took the store from a diminishing 540 top-ten rankings to 639 in six weeks. Traffic followed. By every standard in this industry, that chart was the happy ending, and that is where every case study you have ever read stops.

Act two: the decay. The rewrite was treated the way most stores treat optimization: as a project. Done, shipped, finished. Over the following months the store slid from 639 to 481 top-ten rankings — below where it started. The work did not merely stop paying. The position eroded past the original baseline while nobody was touching it.

Act three: the recovery. In February 2026 the store moved to weekly optimization — a schedule, not a project. Six months later, total ranking keywords were up 32 percent, the first sustained climb in three years, with top-ten rankings recovered to 622 and climbing.

One store is a data point, not proof. I want to say that plainly before drawing any conclusion, because overclaiming is the disease this industry keeps re-catching. But the shape of that chart — up, down through the starting line, up again under discipline — is exactly what the wider research on content decay predicts. So let me put the outside evidence next to it.

What the research says about content decay

Rankings are getting harder to hold, not easier. Ahrefs' long-running ranking study found that only 5.7 percent of pages reached a top-ten ranking within a year; their 2025 update dropped that figure to 1.74 percent, and found the average number-one result is now roughly five years old. Getting there is rare. Semrush's thirteen-month study measured the other half: of domains that reached the top ten, only 19 percent still held that position seven months later. The turnover is the point. A ranking is not a possession. It is a position in a race that does not pause.

Content decays on a measurable curve. Animalz measured organic traffic to aging content declining at roughly 1.2 percent per week. HubSpot's own historical-optimization program — updating old content instead of only writing new — lifted organic views of updated posts by an average of 106 percent. Both are first-party datasets and I weight them accordingly, but they agree with each other and with our chart: content is a perishable asset, and refreshing it is worth roughly as much as creating it.

The volatility is accelerating in the AI era. Analyses of recent Google updates found nearly 15 percent of top-ten pages falling out of the top hundred after the December 2025 core update, rising to 24 percent after the March 2026 spam and core updates, where over 90 percent of top-ten URLs moved. And the new surfaces amplify the stakes: in Ahrefs' 2025 analysis, AI Overviews cut clicks to the top organic result by about a third.

AI answers prefer fresh pages. Ahrefs analyzed seventeen million AI citations and found cited content averages 25.7 percent fresher than what ranks organically; ChatGPT's citations ran roughly 400 days newer than organic results for the same queries. In plain terms: the newest search surfaces have the shortest memories. A page optimized once, years ago, is precisely the page an AI engine passes over.

Commerce platforms enforce decay by rule. Google Merchant Center expires product data thirty days after its last refresh. Not as a suggestion — as policy. Stale catalogs literally fall off shopping surfaces. And for the majority of stores still carrying manufacturer descriptions, Google's John Mueller has been blunt: no penalty, but for a generic query Google shows one page from among the duplicates and filters the rest. Fifty stores share the supplier's copy; one survives the deduplication; the other forty-nine are invisible through no individual fault.

Why the industry keeps selling projects anyway

If decay is measurable and predictable, why does nearly every SEO engagement end with a deliverable and a handshake? Partly because projects are easy to sell and easy to invoice. The results of that model are also measured: an industry-wide survey of twelve hundred small-business owners found SEO services scoring a Net Promoter Score of zero, with 65 percent of owners having churned through multiple providers. The dissatisfaction concentrates among businesses spending under 500 dollars a month — the exact tier where "we did the project" goes to die quietly while the retainer continues.

The stores were not wrong to want the project. They were sold the wrong shape of work.

The discipline conclusion

Here is the cash version of the argument. Industry benchmarks commonly put eCommerce advertising spend around ten percent of revenue. The store in the chart above has run at 3.2 percent for three years — on the order of seventy thousand dollars not spent on ads, on a business doing about 350 thousand a year — because organic held the load. That is what the discipline is worth in money an operator can verify.

Optimization is not a project with an end date. It is a discipline with a schedule — the same way inventory, bookkeeping, and reordering are disciplines. It is also the maintenance half of Visibility Optimization: one foundation, kept current, serving every surface a shopper might ask. The chart's three acts say it, the ranking-longevity studies say it, the freshness data from the AI engines says it, and the platforms enforce it by policy.

The work never stopped mattering. It just stopped being optional to keep doing it.

Sources and dates

External figures, each linked above and dated here: Ahrefs ranking-longevity study (2017, updated May 2025); Ahrefs AI-citation freshness analysis (17M citations, 2025); Ahrefs AI Overviews click-through study (April 2025; Ahrefs has since published a 2026 re-run reporting a steeper decline); Semrush's thirteen-month ranking study (28,000 domains, 2021–2022, published 2023); Animalz content-decay measurement (2018 analysis, page maintained since); HubSpot's historical-optimization program; SE Ranking's core-update analyses, December 2025 and March 2026 (the March figures cover that month's spam and core updates together, which SE Ranking notes cannot be separated); Google Merchant Center's data-expiration policy; John Mueller on duplicate product descriptions, as reported by Search Engine Journal; and the Backlinko/Northstar survey of 1,200 small-business owners (SEO services NPS of zero; fieldwork 2019, page updated 2026). Store figures are ScandinavianShoppe.com's, published with the owner's permission.