This blog has been live for three days. That is not a typo — three days. And already there are things I did not expect, problems I had to fix, and actual data to look at. Here is the unfiltered recap of my first few days as a blogger.
The Numbers (So Far)
I asked Claude to pull up both the Google Analytics data and the raw server logs. Here is what we found:
- 45 real visitors according to Google Analytics
- 22 from the United States — the primary target audience
- 5 from Poland, 4 from Korea, 2 from UK, 1 from Germany
- Googlebot crawled the site 524 times — and the number is increasing daily (81 on day 1, 388 on day 3)
- Most visited pages: Homepage (41 views), About (3), first blog post (3)
Is 45 visitors in three days a lot? No. But for a brand new blog with zero promotion, no social media sharing, and no backlinks, it is a sign that things are working. Google is finding the site, crawling it aggressively, and real humans are actually reading posts.
The fact that most visitors came from the US tells me the LA server location and English content strategy are on the right track.
Thing That Broke: Scheduled Posts Did Not Publish
This was the biggest headache so far. I had carefully scheduled posts to publish twice daily — 9 AM and 3 PM Los Angeles time. But when I checked the WordPress dashboard, two posts showed “Missed schedule” in red instead of being published.
I screenshotted the dashboard and sent it to Claude. It diagnosed the problem quickly: WordPress relies on its own internal cron system, which only runs when someone visits the site. With almost no traffic, nobody was visiting, so the cron never triggered, and the scheduled posts never published.
Claude fixed it by setting up a system-level cron job that hits the WordPress cron endpoint every 5 minutes, regardless of whether anyone visits. Problem solved — except then two of the fixed posts reverted back to “scheduled” status. Turned out the GMT timestamps were not set correctly. Claude had to update those manually too.
This is the kind of thing that would have taken me an hour to debug on my own. With Claude, it was about 10 minutes of back-and-forth screenshots.
Thing That Broke: The Theme Had No Titles
When I switched from the default theme to Astra, the blog posts showed up as nice cards with images — but the titles were completely missing. I documented the full story in my theme redesign post, but the short version is: Claude tried five different approaches before finding one that worked. It was a good reminder that AI does not always get it right on the first try — but it keeps trying until it does.
Thing I Did Not Expect: Bots. So Many Bots.
When I first asked Claude to check server traffic, I got excited — thousands of hits! Then Claude filtered the data more carefully and explained that the vast majority were bots, crawlers, and automated scanners. Some were useful (Googlebot indexing my content), most were just noise (vulnerability scanners probing for exposed config files).
This was a useful reality check. Server logs tell you how many requests your server handled. Google Analytics tells you how many actual humans visited. The difference between 3,000+ server hits and 45 real visitors is mostly bots.
Thing That Went Right: Google Is Paying Attention
The most encouraging data point is the Googlebot crawling pattern. On day 1, it crawled 81 pages. By day 3, it was crawling 388 pages in a single day. It is reading my posts, my About page, my sitemap, even my category pages.
This means Google is treating the site as active and worth indexing. The more frequently Google crawls, the faster new posts show up in search results. I am not ranking for anything yet (way too early), but the foundation is being built.
What I Would Do Differently
- Set up system cron from day one. The missed schedule issue was completely avoidable. If you are running WordPress on a VPS with low traffic, always use a system cron instead of relying on WordPress’s built-in cron.
- Change the theme before publishing. I published five posts with the default theme and then changed it. Should have done the theme first.
- Set GMT timestamps correctly. This caused two posts to get stuck in “scheduled” status even though their time had passed. A small technical detail that caused a real headache.
What Is Next
The immediate plan:
- Keep publishing on schedule (now fixed with system cron)
- Reach 20+ posts before mid-June
- Apply for Google AdSense once the site has enough content and age
- Build the Telegram bot for mobile blog management (coming soon)
- Check back in a week with more traffic data
Three days in and I have already learned more about WordPress, SEO, and web hosting than I expected. Most of that learning happened through problems — things breaking and getting fixed. Which, honestly, is how most real learning works.
How This Post Was Made
I told Claude: “Write a first-week recap. Use the actual traffic data we just looked at. Include the things that broke.” Claude pulled the analytics numbers from our earlier conversation (where I had sent a screenshot of the Google Analytics dashboard) and the server log analysis it had run. The mistakes and fixes are all real — they happened in our conversation over the past three days.
I specifically asked Claude to include the “What I Would Do Differently” section because I think hindsight is one of the most valuable things a blog like this can offer. It is easy to write about what went right. The useful stuff is what went wrong.
This post was written with Claude AI. I provided the direction, topic, and key points in Korean — Claude turned it into the article you just read.