A sitemap is a map you draw for Google. Mine sat folded and unread for three months.
On July the 6th this year, I did something that felt responsible. I opened Google Search Console, found the place where you submit a sitemap, typed in the address of my store's sitemap, and clicked submit. A tick appeared. The interface said everything was fine. I closed the laptop feeling like a proper webmaster, and I did not look at that screen properly again for over three months.
This is the story of how that one click quietly buried months of my work, how I found out, and what the numbers looked like once the fix landed. It is not a dramatic story. Nothing crashed. Nothing got hacked. That is exactly what makes it worth writing down, because the failures that hurt small businesses are almost never crashes. They are the silent ones. The filed-away ones. The "set and forget" ones that turn out to be just "forget."
What a sitemap actually is, in plain words
Let me explain this the way I wish somebody had explained it to me. Google does not automatically know about every page on your website. It crawls, it follows links, it guesses. A sitemap is you handing Google a neat list: here are all my pages, please come and read them. For a small site like mine, it is the difference between hoping Google finds your shop and handing out directions.
When I submitted my sitemap in July, the system accepted it and reported zero discovered pages. I remember noticing that and, shamefully, not chasing it. I assumed the numbers took time. Some things do. This is the trap. In online business, every real signal is wrapped in so much waiting that a genuine warning looks exactly like normal patience.
The truth: I had submitted the wrong map
Three months later, during the October audit of my sites, my AI assistant went looking for why my pages were not getting indexed and found the truth in minutes. The address I had submitted in July, /sitemap.xml, was not my store's real sitemap at all. The SEO plugin that runs my store, All in One SEO, had changed how it builds sitemaps in a newer version. The real one now lives at /sitemap_index.xml, a master map with five smaller maps inside it: posts, pages, products, categories and tags. The old address I submitted pointed at something Google could not read as a list of pages. It read it as an index with nothing inside. Zero pages discovered. Forever.
Meanwhile, my robots.txt file, the signpost that tells crawlers where to look, still pointed at the old address too. So Google had been politely directed to a locked door, three months in a row, while I kept writing new posts and wondering why the visitors never came.
Here is the part I keep replaying. Every post I published in July, August and September was handed to Google through a door I had personally, carefully, clicked shut. The content was fine. The images were original. The internal links worked. The pages simply never got invited to the party.
The fix: ten minutes that should have happened in July
The correction itself was almost embarrassing in its simplicity. Submit the correct sitemap address, /sitemap_index.xml, in Search Console. Done. Google accepted it and began crawling properly, and over the following days the discovered-page count climbed past four hundred. That is the number that should have been climbing all along: 416 pages indexed, out of the 600-odd URLs on the store. All that content, waiting behind a typo I never checked.
I also learned the language of fresh submissions. When you first submit a sitemap, Search Console sometimes says it could not fetch it. My stomach dropped when I read that, thinking I had broken something new. It turns out that is normal for a fresh submission; Google simply rereads it on its next schedule. If you are starting out and you see that message on day one, do not panic like I did. Panic later, only if it still says that in a week.
The diary blog had the same disease
Because the universe has a sense of humour, the audit found the same illness on this very diary. Search Console showed that of my roughly fifty diary entries, exactly one was indexed. One. The rest sat in a state called "discovered, not indexed," which is Google's polite way of saying "we know your pages exist, we have decided they are not worth reading yet." I write two to three of these entries a week. I have been writing them into a void with slightly better manners.
Then it got worse. In the middle of October, the Search Console property for this diary vanished from my account entirely. Not deleted by me, not paused. Gone. The daily indexing routine had been quietly failing for days while reporting that everything was fine. We rebuilt the property, verified it again through my Analytics connection, and the indexing requests started flowing the same morning. The lesson there was the same lesson wearing a different coat: systems that report silently fail silently.
What the numbers look like now
I want to be honest about what a fixed sitemap does and does not do, because I dislike the gurus who sell every fix as a rocket. Submitting the correct map did not flood my store with customers. The diary still has one or two indexed entries this week, not fifty. Sitemaps do not convince Google your content is good. They just make sure Google can find the argument in the first place. The judgment still has to be won the slow way, post by post, with original writing and real experience.
What changed is the direction of the numbers. Before the fix, the store was invisible by accident and I did not know it. Now the pages are discovered, the daily indexing requests go out on schedule, and when a post takes off it will take off for real reasons instead of being capped by a mistake from July. You cannot control whether Google likes your work. You can make sure your work is at least on the table.
The five-minute check I now do every week
Because I promised practical, here is the exact check, written for the version of me from July who did not know any of this. If you run a website, open Google Search Console and look at the Sitemaps page. You should see, next to your submitted address, a number of discovered pages. If that number is zero, or obviously lower than the number of pages you have published, you have my problem. If it says "couldn't fetch" for a fresh submission, wait a week before worrying. Then find your real sitemap by checking your SEO plugin's settings, because most of them generate it for you at an address like /sitemap_index.xml or /sitemap_index.xml style paths, and make sure what you submitted matches what exists.
Second, if you use WordPress, look at your robots.txt. Type your domain plus /robots.txt into a browser and read the line that says "Sitemap." If that address is dead, your signpost is pointing at a locked door, exactly like mine was. Five minutes. I lost three months to it, and I only found it because an AI I pay in credits went looking on a Tuesday.
Set and forget is just forget
So here is my rule now, and it is the reason this entry exists. Nothing on my websites is "set and forget" anymore. Everything that was set once gets checked on a schedule: the sitemap, the robots.txt, the ads.txt file, the subscribe forms, the affiliate links. My AI assistant runs most of these checks daily now and reads me anything that has changed. The July version of me submitted a map and trusted a tick mark. The October version of me asks, every week, what the machine actually sees.
If you run a website and you have not looked at your sitemap status in a while, this is your nudge. Open Search Console. Check how many pages Google has actually discovered, not how many you assume it has. The check costs you five minutes. My three months of quiet invisibility cost me more than that, and the worst part is that I paid the bill without ever seeing the invoice.
Related entries from the journey
This is not the first time the invisible machinery of the web has humbled me. I wrote about six months of writing into the void and how SEO actually works, and about learning to write for Google instead of myself. And if you want the other side of the same coin, read what happened when AdSense looked at my site and said insufficient content.
The check I should have run in July
I finally ran the check properly, and the result is both a relief and an embarrassment. Every one of the 61 URLs in my sitemap returns a clean 200 today. Not one redirect, not one timeout, not one error page. I checked all of them, not a sample, because a sample is exactly how you miss the one bad entry that teaches the crawler to distrust the whole file.
The embarrassing part is what Google still believes. My Pages report says seven indexed out of 61, and that number is frozen at 23 October, so it is almost two weeks stale. A live query on the site itself returns fourteen URLs, twelve posts plus the homepage and the About page. Sitting underneath that are eighteen posts tagged as redirect errors, with last crawl attempts spread between 14 September and 6 October. Every one of those attempts happened before the fix. The restarted validation has processed none of them in ten days.
Which means the dead sitemap in this story is not the only dead thing. The validation queue is dead too, quietly, and it does not send you an email to say so.
What I do about it now, in order. Press Validate fix rather than assuming it is still running, because ten days with zero processed is not patience, it is a stalled job. Resubmit the sitemap so the crawler gets a fresh timestamp to work from. Then use the daily inspection quota on the posts that matter most, roughly a dozen a day, oldest and emptiest first. And stop reading the Pages report as if it were live data, because it is a photograph, not a window.
The lesson I keep relearning on this site: a fix is not finished when the thing works. It is finished when the crawler agrees that it works.
And a note for anyone reading this with the same symptom: an old post that will not index is almost never a content problem. Check the status code first, then check what host the sitemap advertises, then check whether validation is actually moving. Mine was behind a mobile 404 for three months and no amount of rewriting a single paragraph was ever going to fix that.
A sitemap nobody crawls is a letter nobody opens. Post it, then check the reply.
If you take one action from this post, take the boring one: open your sitemap, list every URL, and check the status of each. Sixty one entries took me nine minutes. Nine minutes would have saved three months.
Newest entries from the diary
Written by Justice Manyika — Johannesburg, South Africa. This diary records my real journey, in my own words.
If this helped, follow the journey — drop your email in the box below and every new entry lands in your inbox. No spam, unsubscribe anytime.
Comments
Post a Comment