[Monthly Report] Growth Record of note Operations (April 2026 Issue)

It has already been one month since I started note.

From my first post on March 30th to today, April 27th, I have published 24 articles, reached 101 followers (Thank you for surpassing 100 followers!), and exceeded 1,000 total views. While the numbers alone might make it look like things are going smoothly, looking back at them with a CS perspective, there were several unexpected trends. I am writing this to organize those thoughts.

The biggest discoveries this month were two things: "the 'Like' rate (NPS for writers) is far exceeding industry benchmarks" and "the underlying strength of past articles was stronger than I thought."

This month's figures (as of April 27th)

Since this is the first issue, I cannot compare it with "last month-end," so instead, I am writing based on April 3rd, when I started tracking.

The thing that caught my attention most when looking at the numbers is that the engagement rate (Likes ÷ Views) has continued to exceed 30%.

Just like with CS NPS or CSAT, the direction of the movement is more important than the absolute value, but it is a bit strange that this level is continuing. I feel that the structure of the note follower base (followers basically give Likes) is also having an effect. Even so, compared to the "industry benchmark of 5-10%" often mentioned online, it is a genuinely high figure.

Top 3 articles that grew this month

1st place: The difference between Customer Support and Customer Success, a story about not being able to answer when asked in an interview (80 views / 23 Likes / Like rate 28.8%)

An article that resonated with job changers thinking about a career change to CS. I think the specific failure episode of "stumbling in an interview" was more attractive than abstract CS theory. In terms of a CS help center, it feels exactly like the movement of people searching and arriving at "top FAQ articles" while having specific worries.

2nd place: [Self-introduction] Nice to meet you, I'm "Dan." (74 views / 33 Likes / Like rate 44.6%)

My first post. With a Like rate of 44.6%, it was by far the highest among all articles. In CS terms, it might be close to the phenomenon where "initial customers (users immediately after onboarding) have the most enthusiasm." It is a movement unique to the initial stage where you can have friends and acquaintances read that you "started note." If I published the same article in half a year, I probably wouldn't get these numbers.

3rd place: What does EdTech Customer Success do? The overall picture seen after doing it for over 7 years (68 views / 16 Likes / Like rate 23.5%)

A long-form article summarizing the overall picture of the industry. It is the only one in the Top 3 that deals with a "question that doesn't have an immediate answer (the entire industry)." It is a prime candidate for search traffic, and in CS terms, it plays the role of "educational content for users who don't know the company's product." It is positioned as an article to help readers build a framework in their heads that "EdTech CS is this kind of job."

Disappointments and unexpected trends

What was unexpected was that the "answer key to the story about not being able to answer the 'difference between support and CS' in an interview" gradually increased in views in the latter half of April.

There were even days with +14 views in a single day.

This is paired with the #1 article, 'The Difference Between Customer Support and Customer Success,' and it seems highly likely that readers who read the 'Story of Not Being Able to Answer' became curious about 'So, what is the answer?' and navigated to it.

In CS terms, this is exactly the movement of 'jumping from a help center article to related articles.' I didn't intentionally design the flow, but as a result, it felt like the two articles were read as a set.

What was disappointing was 'The Story of How a Humanities-Major CS Improved Completion Rates with SQL.' Even though the title seemed like a strong hook, it only had 7 likes (12.7%) against 55 views, so engagement didn't grow relatively well. The content was about 'completion rates' and 'trends,' which leaned more toward EdTech CS, so it might have been a bit off from the preferences of the average note reader.

Interpreting through the CS Framework

Among the KPI mappings shown in the framework section, the two that were particularly effective this month were 'Long-tail customers ↔ Views of past articles' and 'NPS ↔ Like rate.'

The long tail was stronger than I expected. In the last week of April, even on days when I published new articles, the majority of views came from past articles. It's the same structure as looking at 'what percentage of a month's ARR is accounted for by existing customers' in CS; for note as well, it feels like the 'underlying strength of past articles' provides more stability to daily views than the 'immediate effect of new articles.' This is a movement that beginners often overlook, and I was a bit surprised when I noticed it myself. New articles are the instantaneous maximum wind speed, while past articles are the base power. I realized that things don't run without both.

The like rate was also at a surprisingly high level. The overall engagement rate remained above 30% throughout April. I see many articles online stating that 'the industry standard for note like rates is 5-10%,' but my numbers were significantly higher than that.
The reason is likely because I am in a highly dense onboarding state. While the number of followers is small (around 100), followers are the core of the readership, and everyone basically likes the posts. As the number grows to 500 or 1,000, new views (inflow from non-followers) should increase, and the like rate should naturally decrease. I believe this is the same movement as 'dilution of active rates' in CS.

Therefore, I interpret this high level not as 'amazing,' but as 'a number unique to this scale.' It should go down as the number of followers increases from next month onwards, but I think I should read that not as a bad sign, but as proof that the base has expanded.

Hypothesis → Verification → Learning

Things I actually tried this month (since it's the first time, there are no past hypotheses)

I tried two things in the second half of April. One was 'whether changing the title would move views on dormant articles,' and the other was 'whether serialization affects readers' continued following.'

Verification Results

The title change (changed to 'The Story of Getting Ahead with a Part-time Job Without Waiting for the Recruitment Exam') resulted in +9 views in 3 days after the change. Compared to the previous daily average of +0.5 views, it clearly moved.
Serialization (publishing the framework section) resulted in +4 followers in 26 hours after publication. It accelerated from the usual pace of +2.9/day. I think the set effect of X notification + magazine publication also played a role.

Learning

Both 'moved,' but the amount of movement was still small. Since the sample size is insufficient, I feel it's a bit weak to say definitively that it was 'effective.' Next month, I want to consciously repeat the same experiment to confirm reproducibility.

Even in CS work, there is a phase of checking reproducibility across multiple companies before rolling out a change that occurred at one company to the entire company. I intend to move in the same way.

Hypothesis for Next Month

Hypothesis 1: Will the publication of the 3rd installment of the series (Observation Section) cause a secondary rise in views for the Framework Section?

How to look at it: After the Observation Section is published, I will track the view trends of the Framework Section.
Prediction: It will be read by looking back from the Observation Section, and there should be a ripple effect of about +20% on the views of the first installment.

Hypothesis 2: If the number of magazine subscribers increases, will the initial views of new posts increase?

How to view it: Record the weekly trend of magazine subscribers and compare the 24-hour views of newly published articles.
Prediction: Magazine subscribers should become regular readers, stabilizing the initial performance.
I will check the results in next month's report.

Conclusion

I think that for note operations, just like in CS work, it is more meaningful to look at 'how we moved compared to ourselves last month' rather than 'how we compare to the industry average.' Since this is the first time, the only comparison was the starting point, but from next month, I will be able to write in terms of 'month-over-month' changes. I'm looking forward to that a little bit.

What kind of trends did you see on your own note this month? I would be happy if you could share in the comments things like 'I am looking at these numbers' or 'I am testing these hypotheses.' I am still experimenting myself, so I would love to hear your perspectives.

The next installment in the series is Observation Edition (S1-3). This will be the episode where I reveal the mechanism of 'how I collect these numbers and how I analyze them with AI.' The numbers used in this report were also all collected using that mechanism, so I will structure it so you can read it as a way to check the results.

Hypothesis → Verification → Learning. I will continue to cycle through this next month as well.