🎧 Listen to the Discussion(two hosts talk through their thoughts on this article)
This is an AI-generated audio overview of this article, created with Google NotebookLM. Two hosts discuss the article together, sharing their own thoughts and commentary — it is not a reading of the article text.
After retiring at 65, I decided to rebuild this blog properly. The part that gave me the most trouble, by a wide margin, was getting approved for Google AdSense.
Let me start with the outcome: I was rejected three times before approval. Each rejection email was word-for-word identical to the last, and none of them told me what was actually wrong. I spent several months fixing things in the dark, with no face and no reason on the other side of the screen.
I am writing this down in case it helps someone who is stuck in the same place. Here is what I did, and where I wasted my time.
What was actually wrong
The first rejection said "low value content." I had more than 150 articles. I was annoyed, and a little ashamed.
Working through it with an AI, I found two causes.
First, my articles had no URLs of their own.
Back then, clicking an article opened it in a modal window. It looked modern, but the URL stayed as something like /?article=64. There were no independent pages. To Google, my site was a single thin page. What my readers could see, the machines could not.
Second, I was duplicating my own old site.
My previous WordPress site was still live and still indexed. The same articles existed on two domains, and Google had no way to tell which one was the original.
So I had submitted a site that looked thin and looked like a copy of someone else. No wonder it failed.
Each article getting a light of its own, one stepping stone at a time, out of the fog it had been hidden in behind a modal window. Light the nearest stone first, and the next step becomes clear.
What I fixed, with AI alongside me
I am not a programmer. I got through this by talking to Claude and moving one step at a time.
Every article got its own page
I regenerated all 154 articles as static pages in both Japanese and English — 308 pages in total. Each one has its own title, description, canonical tag and hreflang. I kept the modal for convenience, but the cards now link to real URLs.
This was the biggest job of the lot, but once the machinery existed, new pages generate themselves whenever I write. Had I done it by hand, I would still be working on it.
I moved out of the old site properly
I set up 143 redirects from the old article URLs to the new ones, matching them automatically on title and publication date. 98.6% matched on the first pass; I checked the rest by hand.
I did not let the old domain expire. I renewed it for another year. If I had dropped it, every link anyone had ever made to my writing would have gone nowhere.
I rescued 2,771 images
After the move I discovered that every image inside my articles had vanished — they were still pointing at the old server. Fetching them through the domain did not work. Neither did the server name. In the end I pulled them down over FTP, directly. I recovered 2,770 of 2,771. The missing one turned out never to have existed.
I dug lost articles out of the archive
Some articles ended after the introduction. It turned out that years ago I had published the second half on a separate site — one I had since deleted.
I nearly gave up. Then I found the full text preserved in the Wayback Machine. All eight articles came back. I will admit to a small fist-pump.
I added a contact page and a disclaimer
After the third rejection, I went through the whole site again and noticed something.
There was no way to contact me.
There was a link labelled "Contact," but it led to a newsletter signup box. My email address appeared nowhere on the site. Neither a reader nor a reviewer had any way to reach me.
I built a contact page and a disclaimer page, and published an email address on my own domain.
Two traps worth knowing about
The AI's suggestion pointed the wrong way
When I added the email address, the AI first proposed assembling it in JavaScript — so that crawlers could not read it. Technically sound. Standard practice against spam harvesters.
But my entire purpose was to let the reviewer and the crawler see my contact details. The suggestion pointed in exactly the opposite direction.
I wrote the address as plain text in the HTML and used JavaScript only to turn it into a clickable link. AI suggestions are clever. Clever and right for this particular goal are two different things.
It passed in preview and failed in production
That email address displayed as plain text in the preview environment. On the live domain, it had been replaced with a scrambled string.
Cloudflare's "email address obfuscation" feature was rewriting my HTML as it served it. The care I had taken to avoid JavaScript obfuscation was being undone somewhere else entirely. Turning the setting off solved it — and taught me to always fetch the raw HTML from the live site before concluding anything.
My longest detour: starting the account over
This part is specific to my situation, but it may be useful.
I hold Japanese nationality and live in the UK. My AdSense payments profile had been created in Japan, which meant the UK tax treaty could not be applied.
The country on a payments profile cannot be changed. Each country needs its own profile, and there is no option anywhere in the interface to move an existing account onto a different one.
I contacted support, followed their guidance, closed the old account and opened a new one at my UK address. That detour cost me months. I lost my unpaid balance, though thankfully it was small.
If you live abroad and are about to start with AdSense: create the profile in your country of residence from day one. Correcting it later is far harder than it sounds.
What I learned, at 66
This may be more useful than the technical parts.
Check your fears against facts. "Health articles are YMYL, so they'll never get indexed." "Maybe the ad code isn't installed." Worries arrive freely. Every one of mine turned out to be wrong. The unindexed articles all went through within minutes of requesting indexing, and the ad code was exactly where it should have been. Fix things on a hunch and you break what already worked.
The dashboard wording tells you nothing. The "comply with program policies" checkbox is a self-declaration, not a review item. Nowhere does it say what the actual problem is. Rather than exhausting yourself trying to read words that were never written, decide on your own inspection list and work through it.
Investigate, test on one or two, then apply to everything. Rewrite 308 pages at once and get it wrong, and undoing it will eat your entire day. Once I started keeping to that order, my rework dropped away almost completely.
Stop, and rest. This might matter most of all. My concentration and memory are not what they were. After hours glued to a screen, I can feel my judgement slipping. I once mistook an old screenshot for a current one and lost half a day chasing a problem that no longer existed.
Building a rhythm between effort and rest is a healthy-ageing question. It is also, quite directly, a question of work quality.
Closing
When the approval email finally arrived, I did not leap out of my chair. What I felt was a quieter kind of relief — that I had reached the starting line at last.
Honestly, the ads themselves matter less to me than what those months did to the site. Every article has a real URL now. The lost images are back. The truncated pieces are whole. There is a way to reach me. I can finally look my readers in the eye.
An AdSense rejection is an infuriating thing to receive. But looking back, it was also a fair piece of homework: go and rebuild your site properly.
If you are sitting in front of that same screen right now, sighing — a rejection with no reason attached is not a verdict on your writing. It may simply mean nobody can see it yet. Make it visible, and the road opens.
This is a record of my own experience. It is not a guarantee that following the same steps will get you approved. AdSense criteria and policies change, so please check the current official documentation.
Other books you might like
You're welcome to take a look.
EAST TO WEST: The Heart of Haiku, the Light of Scripture
Thinking about how the unseen became visible in this story makes me want to open this book of quiet light again, where haiku and scripture meet. This was my father-in-law's final work.
View on AmazonCataract Surgery-A Memoir: Double cataract surgery in Osaka, Japan
Another record of mine, at 66, of feeling my way through a cause that was hard to see at first.
View on AmazonSushi Business in Britain: the complete four-volume series
A novel series following thirty-five years in the British sushi trade, from apprentice to factory manager to a small sushi school.
View on Amazon