SuzuLabo

English homeCase studies

Evidence · Measurement conditions · Limits

Case studies

We are a small Japanese company, and much of our work is done under agency contracts or as ghostwriting, which means we cannot publish it. Rather than fill this page with logos we have no right to use, here is what we can show, with the measurement conditions attached.

Three things worth reading first

  • We separate first-party data from third-party estimates. SEO tool figures are estimates produced by a third party's model, not analytics from the client's own account. We label which is which every time.
  • We do not claim causation we cannot demonstrate. A page that grew after we worked on it is not proof that our work caused the growth. Search results, competitors and the client's own activity all move at the same time.
  • These are Japanese-market results. They show that we can execute Japanese SEO, content and web work. Where a client was a Japanese company rather than an overseas company entering Japan, we say so.

Case 01 · Japanese SEO content

Canva Japan — paid SEO content for the official blog

ClientCanva Japan
WorkJapanese SEO content production
PublishedOn Canva Japan's official blog
StatusCompleted assignment

What we did

SuzuLabo was engaged for a paid SEO content assignment for Canva Japan's official blog. Our scope covered the article production process:

  • Japanese keyword research and topic selection
  • The article outline, approved before writing
  • The Japanese first draft
  • Supporting image creation for the article

The final article was published after Canva's own editorial review.

Screenshot of the published article on Canva Japan's official blog, showing the breadcrumb, the Japanese article title about sole proprietorship, a summary and the article's hero image.
The published article on Canva Japan's official blog. Captured 5 August 2026. The page, its design and the Canva brand are Canva's own; this shows where the work was published, and is not a partnership, sponsorship or endorsement. Our scope was the production process listed above, and the article was published after Canva's editorial review.

What happened afterwards

A third-party SEO tool estimates that the published URL peaked above 1,000 monthly organic visits in April 2026, measured across September 2025 to August 2026.

Line chart of estimated monthly organic visits to the published Canva article from September 2025 to August 2026, rising through early 2026, peaking above 1,000 in April 2026 and declining afterwards.
Estimated monthly organic visits, September 2025 to August 2026. Third-party SEO tool estimate, not first-party Canva analytics, and it does not establish causation. The chart labels are in Japanese: the vertical axis is average organic traffic and the horizontal axis runs from September 2025 to August 2026. The decline after April is part of the picture, so we are showing it.

How to read that number. It is an estimate from a third-party SEO tool, not Canva's own analytics, and we do not have access to Canva's analytics. Canva's blog is a large, established property with its own authority, promotion and internal linking, and we contributed one article's production process within it. We are showing the outcome of an article we worked on, not claiming that our work alone produced the traffic.

Why we show it

Because it is verifiable. The article is public, the assignment was paid, and the scope above is what we actually did — not a longer version of it.

View the published article

Case 02 · Japan-market SEO programme

Japan-market SEO for a Japanese B2B training provider

ClientName withheld
MarketJapan
WorkJapanese SEO strategy and article production
Programme sizeRoughly 20 articles

What we did

A Japanese-language SEO content programme for a B2B training provider: keyword and intent research, topic selection, article production and on-page work, run as a programme rather than as one-off articles.

Results

As of July 2026, third-party SEO tool estimates showed:

#1for the main target topic
319estimated monthly visits from one article
28search queries the article appears for
≈20articles in the programme

How to read those numbers. They are third-party SEO tool estimates as of July 2026, not the client's own analytics. Rankings and traffic move over time, and these figures will not be identical when you read this.

Why we show it, and what it does not prove

This was a Japanese company selling in Japan, not an overseas company entering Japan. It demonstrates that we can research, plan and produce Japanese SEO content that ranks. It is not evidence of a track record with overseas clients, and we would rather say that plainly than let an anonymized case imply something it does not support.

Case 03 · Our own work

Our own site, and our own measurement

The one project we can show you in full, without anyone's permission, is our own.

A bilingual site we built and run

suzulabo.co.jp is our Japanese site. The English section you are reading now is part of the same property: separate English content rather than machine translation, its own language annotations and canonical setup, structured company data, and a single inquiry form with English validation and confirmation handling.

Japanese · suzulabo.co.jp Screenshot of the Japanese SuzuLabo homepage, with a Japanese headline set in a serif typeface, a subheading and two calls to action.
English · suzulabo.co.jp/en/ Screenshot of the English SuzuLabo page, with a different headline, a fixed-price offer card and its own calls to action.
Two versions that are not translations of each other. Different headlines, different proof, different first action, because the two readers arrive with different questions. Captured 5 August 2026.

If you want to see how we treat Japanese typography, company information, form fields and the path from a page to an inquiry, the fastest review is to open both versions and compare them.

Measuring AI search visibility on ourselves first

Before we sell AI search visibility work, we run the measurement on our own brand.

Method. On 5 August 2026 we ran six purchase-decision questions across five AI systems — ChatGPT, Gemini, Perplexity, Claude and Microsoft Copilot — for 30 answers in total. The questions were asked in Japanese, in new or temporary conversations, with web search enabled, without personalization history. We saved the full text of every answer, the cited links, and a full-screen screenshot of each, and we counted the answers that did not mention us.

6 purchase-decision questions × 5 AI systems: ChatGPT, Gemini, Perplexity, Claude, Copilot = 30 answers, including the ones that ignore us SAVED AS EVIDENCE Full answer text Cited links Full-screen screenshot Conditions recorded every time: language, new or temporary conversation, web search enabled, no personalization history, and the order the systems were run in — so the same test can be repeated after changes and compared honestly. Run on our own brand in Japanese on 5 August 2026. A single run is a baseline to re-measure against, not a score.

What that is, and is not. It is a starting measurement of our own brand in Japanese, taken before the work — not a claim about what we can achieve for a client, and not a market study. Answers vary between systems, between phrasings of the same question and between runs, so a single measurement is a baseline to re-measure against, not a score. We run it on ourselves first because a vendor who has not measured their own starting position is not in a position to interpret yours.

Where this goes next. The same measurement, in English, for the questions an overseas company would actually ask when looking for help entering Japan, is the next step. We will publish the method and the results the same way, including the questions where we do not appear.

What we cannot show you yet

Being straightforward about this is part of the service.

  • Agency and white-label work. A substantial share of our SEO and content work was produced under other companies' names. We cannot publish it, and we will not describe it in a way designed to let you guess.
  • Work under NDA or without written permission. Where a client has not given written permission to be named or described, we do not publish the work — anonymized or otherwise. Removing a company name does not remove a confidentiality obligation.
  • Bilingual website production for an overseas-facing project. We have relevant production experience that we are not publishing here because the written permission to describe it is not in place.

If you need references before committing to a paid engagement, ask. We will tell you what we can arrange and what we cannot, rather than producing testimonials we cannot substantiate.

How we will report your project

The same rules apply to your work as to the cases above.

  • We agree in advance what may be published: your name, your industry only, your metrics, screenshots, or nothing at all. These are separate permissions, and any of them can be declined without affecting the work.
  • Reported figures carry their source, the measurement period, the metric definition and the date they were taken.
  • We report what did not move alongside what did.
  • Nothing is published as a case study without your written permission.

Start with a written review

Send your site and what you need to happen in Japan. We will tell you whether we can help, what we would do first, and what it would cost. That reply is free.

Services: Japan SEO · Japanese website localization