Omnilude-tools After Its First Month: What Works and What Still Does Not

Opening#

As I mentioned in my earlier post Introducing Omnilude-tools, I wanted to share the metrics that appeared after the site went live.

As I wrote before, this site was more of an experiment than a project I had high expectations for, and that is still true now. Because of that, I can probably look at the data with a drier and more objective perspective.

Let’s Start With the Numbers#

The numbers used in this post are based on the GA4 overview report and the Search Console page/query/country/device reports.

These numbers are not impressive on their own. Still, compared with the state of simply putting the tools out there a month ago, they are enough to confirm that this is no longer a zero-response project.

Even without marketing, it was possible to confirm that SEO alone can create a certain amount of traffic.

The Most-Viewed Screens on the Site Were Different From the Pages Clicked in Search#

Based on GA4 page views, pages like the home page, Subtitle Converter, Timer, Alarm, and JSON Editor were opened relatively often.

The first thing that stands out is that on-site page views and search clicks are not telling the same story.

For example, Subtitle Converter looks strong in page views, but it does not stand out among the top search-clicked pages. On the other hand, some multilingual pages are reacting first through search impressions and clicks rather than through on-site browsing.

I read that difference like this.

In other words, you have to look separately at the screens people open often and the screens search picks up first if you want the next action to be clear.

Search Surprisingly Reacted to Multilingual Pages First#

If you look at the top pages by search clicks, the picture is a little different from what I first expected.

Two things stand out the most.

First, clicks were not concentrated only on Korean pages. Pages like the French GPS Coordinates Converter, the English Keyboard Converter, the English Alarm, and the Spanish Flashlight reacted earlier than I expected.

Second, impressions and clicks were still separate problems. The English Barcode Generator recorded 613 impressions but only 2 clicks. That single page accounted for 21.9% of total search impressions, yet its CTR stayed at 0.33%.

In other words, the multilingual SEO structure itself is already being indexed fairly broadly by search engines. But there is still a large gap between being indexed and being clicked.

Search Clicks Started Rising Gradually in Late February#

If you look at the daily trend of search clicks during this period, the later half matters more than the beginning.

It becomes clearer when you split the period in half.

In other words, 74.4% of total search clicks happened in the latter 15 days. I would not read this as a huge success signal. It is better read as a sign that search response is gradually appearing over time.

In particular, clicks were relatively higher on 2026-02-18, 2026-02-19, 2026-02-21, 2026-02-23, and 2026-02-26.

Daily clicks still fluctuate, but at least it confirms that this was not just a brief burst of traffic that ended immediately after launch.

The Queries Make the Nature of the Tools Clearer#

The top queries are still small, but they do show, to some degree, what kinds of problems people are trying to solve.

Representative queries included json 파일 수정, online alarm clock, convertisseur de coordonnées gps, keyboard layout translator, 위도경도 좌표변환, and 저항 색띠.

One thing is clear from those queries. What is reacting now is not brand search but problem-solving search with very explicit tool intent. I think that is actually a natural pattern in the early stage of a multi-tool product.

This Is What It Looked Like by Country and Device#

Country data shows that impressions and clicks were not limited to Korea alone.

The interesting comparison here is Korea and the United States. Impressions were almost the same at 590 for Korea and 594 for the United States, but clicks were 12 for Korea and 5 for the United States. In other words, English-language exposure opened up faster than I expected, but click conversion is still weak.

By device, desktop accounted for a much larger share.

74.4% of search clicks came from desktop. My interpretation is that the tools reacting right now, such as developer tools, converters, and editors, are simply more comfortable to use on desktop. This is still my inference, but it is grounded in the data.

What Is Working at This Stage and What Is Not Yet#

Now I can summarize the data a little more directly.

What Is Working#

• Tools with a clearly defined problem are starting to show up in search even when the language changes.
• Pages with clear intent, such as GPS, Keyboard Converter, and JSON Editor, are turning into clicks.
• Search clicks did not end as a one-day spike. They continued into late February.
• The number of pages, queries, and countries generating clicks is increasing, and a small long tail is starting to form.

What Is Not Yet Working#

• The overall scale is still small. 39 clicks / 2,800 impressions / CTR 1.39% is not something I can frame optimistically.
• Many multilingual pages get impressions, but their click conversion is weak.
• The screens that get viewed a lot inside the site and the pages that get clicked first in search are quite different.
• Individual tool pages are reacting earlier than the home page, so the product still depends more on problem-solving entry points than on brand-driven entry.

Things Worth Trying#

This data makes the next actions fairly clear.

• Refine the titles and descriptions of pages like the English Barcode Generator, where impressions are high but CTR is low
• Strengthen pages that already reacted, especially GPS, Keyboard Converter, and JSON Editor type pages
• Realign multilingual page metadata around actual search queries
• Build stronger internal links from the home page to tools that are already getting search response
• Separately review the acquisition path for screens like Subtitle Converter, which get high page views but weak search performance

In short, what is needed now is not to keep adding more tools, but to make the pages already picked up in search more clickable.

Closing#

I do not plan to spend much more time on this tool, but I still think it was worth analyzing once as a deployed service.

What I learned is that if I ever build a more meaningful service, I will need to treat SEO far more seriously. It also leaves me wondering how the numbers might have changed if performance marketing had been run in parallel.

I will come back with something more interesting.

Thank you.