A record of search visibility rising about 25x in 5 months

Tomohiro Iida · Published: July 13, 2026Last updated: October 13, 2026

This report compares five months of Search Console performance after Netsujo migrated its website to Next.js. GSC totals were re-queried directly in October 2026; the GA4 session values are historical figures that have not yet been independently checked.

We compare impressions, clicks, average position and CTR over the same five calendar months, including a lower CTR alongside higher impressions and clicks. This is an n=1 observation from one company website. Multiple changes occurred together, so the chart does not prove that migration or article publication caused the change.

What happened over 5 months (the headline numbers)

The recorded GSC monthly totals show that impressions grew at a faster rate than clicks between February and June 2026. The historical values below use matched calendar months (site=netsujo.jp, dataState=final).

Source audit (October 13, 2026): Google Search Console API, property https://netsujo.jp/, unfiltered Web search and calendar-month dates. All five monthly totals match a separate 150-day daily aggregation. GSC uses America/Los_Angeles date boundaries. The original CSV export was not retrieved. GA4 sessions require separate Analytics permission, and no SEO-causality inference is established.

Verified GSC totals show impressions up about 25x, clicks up 6.8x, and average position changing from 14.9 to 8.2. The earlier article also reported GA4 Organic sessions rising from 108 to 656, but those GA4 figures have not been independently rechecked.

The top clicked queries include branded searches for our company name. The figures above are totals that include branded search, and are not fully separated from the performance of the article content alone.

The old site's one-year plateau

Before the migration, search traffic on the old site was essentially flat. From March 2025 to January 2026, clicks moved up and down within a range of 66 to 123 per month, and average position sometimes got worse (45.7 in June 2025). We believe this reflects a limited page count, with no growing surface of information for search engines to show.

After migrating to Next.js at the end of March 2026 and then starting to publish articles, the way the numbers moved changed. What follows is the story of what we built.

What we built: the migration and 92 articles

At the time of migration, the article count was zero; the first commit in the repository is "Initial commit from Create Next App" on March 27, 2026. From there we built up articles, and as of July 13, 2026 there are 92 static articles in the codebase, and 113 URLs in total under /blog once CMS-managed news and other content is included. The starting point for volume publishing was April 8, 2026, when 11 articles went up at once.

What we built was mostly practical explainer articles on Web3, blockchain, AI, and SEO/AIO. Some of the pages driving impressions and clicks, by cumulative GSC clicks, are below.

• An article on a DID/VC adoption roadmap (60 cumulative clicks, no direct link in the original article).

The RWA guide is one example of why impressions and clicks must be evaluated separately rather than read as the same business outcome.

Comparing the growth rates of impressions and clicks

From March to April, monthly impressions rose from 1,700 to 10,455 (about 6.2x), while clicks rose from 74 to 158 (about 2.1x). These aggregates show different growth rates, not a proven sequence of causation. Query mix and search positions may also have changed.

More impressions do not automatically mean more qualified visitors. Across February to June, impressions rose from 1,296 to 32,890 (about 25.4x), versus clicks from 60 to 410 (about 6.8x). To understand the gap, we need to separate branded from non-branded queries, page groups and ranking bands.

For reference, a rolling 28-day view sometimes shows clicks reaching into the 440s on some days, but that is a different unit of aggregation from a calendar-month total. Because comparing multiples requires matching units, this article uses calendar-month comparisons as the primary figure: 60 in February to 410 in June, about 6.8x.

Interpreting the lower CTR

Across the same months, aggregate CTR declined from 4.63% to 1.25%, a difference of 3.38 percentage points. This is a measured change, not evidence that a particular site improvement caused an adverse effect.

CTR is clicks divided by impressions. Because impressions grew faster than clicks, the aggregate ratio declined. The monthly totals cannot identify why: branded-query share, informational queries and ranking distribution are all possible factors that require separate analysis.

Before changing copy or titles, we should inspect CTR by query intent, ranking band and landing page. Mean position changed from 14.9 to 8.2, but one average cannot identify which individual page should be improved. These are decision criteria, not a general SEO prescription.

How this experience feeds into SIGNAL

Netsujo offers Netsujo SIGNAL, a web sales foundation for the era of search and generative AI. SEO and AIO improvement is one of the elements it diagnoses and improves. This internal record shows why impressions, clicks, CTR and ultimately qualified enquiries need distinct measurement definitions. We apply that distinction when deciding which web improvements to investigate; the data does not establish a universal sequence of outcomes.

SIGNAL does not guarantee search ranking or inclusion in AI answers. What it does is set priorities from real GA4 and Search Console data, and produce implementable deliverables such as copy, structured data, and measurement specifications.

You can start by finding out how your own site looks to search engines and generative AI. This can be checked using public information alone, with no need to share GA4 or Search Console access.