DEV Community: superfire

I Thought My SEO Titles Were Bad. Then I Checked the Positions.

superfire — Wed, 28 Oct 2026 09:51:35 +0000

I run a small site about learning Chinese. Fifty-four pages, all indexed, clean technicals. Last week I pulled 28 days of Search Console data and sat with it for an hour.

The average position was 41.7.

I have read a lot of these posts. Almost none of them show the month where nothing worked. This is that month.

The numbers

Metric	Value
Impressions	490
Clicks	8
CTR	1.63%
Average position	41.7
Queries with any impression	101

And the distribution by position:

Position	Queries	Impressions
1-3	1	1
4-10	4	13
11-20	2	2
21-50	2	3
51+	92	143

Ninety-two of 101 queries sit past page five.

Finding 1: CTR was never the problem

My first instinct was that my titles were bad. So I checked them. The homepage title was 70 characters with the main keywords. The tones page had a hook in it ("Free Test"). Meta descriptions ran 157 to 163 characters with specific claims.

Then I found the counter-example. One page ranks at position 4.3 and converts 33% of its impressions into clicks. Same writing style, same site, same author.

The difference was position, not copy. At position 41, a 1.63% CTR is roughly what you should expect. Position 1 gets somewhere near 25%, but that number belongs to a different situation.

I had been about to rewrite 54 titles. That would have cost a week and changed nothing.

Finding 2: a third of the traffic was the wrong traffic

I grouped the 101 queries by intent. GSC anonymises a chunk of queries, so the visible ones only account for 162 of the 490 impressions, but the pattern was clear enough.

Intent	Queries	Impressions	Share
Looking up a Chinese version of their own name	38	61	37.7%
Learning Chinese	23	43	26.5%
Typos and long-tail fragments	40	58	35.8%

The name-lookup group was things like "chinese name translation" and "convert to chinese name". Those people want to see their name in Chinese characters. They are not learning the language, and they will never buy anything a language site sells.

One page on my site gets 23% of all impressions and produces zero conversions. It is the name generator. It is ranking for the right words and attracting the wrong people.

Finding 3: backlinks did not fix position on their own

I have six dofollow backlinks, from three articles I wrote on dev.to. Two pages carry them. Their average position is 58.9.

A page with no backlinks at all sits at 45.0.

That is not a controlled experiment. Different pages, different competition. Still, it stopped me believing that more links alone would move anything.

What I am doing instead

Not rewriting titles. Not adding more words to pages that already answer their question in the first paragraph.

Two things:

First, making each page's opening line something a machine can quote. Comparing my two versions of the same intro:

Before: "Language is just the surface. Culture is what makes you want to explore deeper."

After: "Chinese culture here covers six areas: food, festivals, calligraphy, kung fu, the 24 solar terms and hanfu. Each section ends with six Mandarin words, 36 in all."

The second one can be lifted and used as an answer. The first one cannot. That is the shape I think AI assistants look for when they decide what to quote.

Second, structured data where it was missing. I audited all 54 pages and found 16 with no JSON-LD at all. Not "no FAQ schema", no structured data of any kind. The culture page was one of them: it had been live for months with nothing. That is a mechanical fix, roughly ten minutes a page.

The honest part

None of this is a growth channel. I am fixing things that should have been right the first time. Traffic will stay small until the site earns enough trust to rank. I do not have a shortcut for that.

If you are at position 40 with 17 impressions a day, stop asking how to improve your CTR. Ask why the page is at 40. Most of the time the site has not earned position 10 yet.

The data behind this post comes from the Search Console API on my own site.

Drafted with AI assistance and reviewed before publishing.

I Built a Free Pinyin Chart, Then Compared It to the Big Names

superfire — Thu, 22 Oct 2026 03:06:17 +0000

I built a free pinyin chart last month. Before I wrote a single line of code, I clicked through every free chart I could find — partly to steal good ideas, partly to make sure I wasn't solving a problem that didn't exist.

Here's what the big ones do well, what they don't, and where mine fits. If you're learning Mandarin, this should save you an afternoon of tab-hopping.

What's already out there

Yoyo Chinese — the most famous, and for good reason. Every sound comes with a short video of a teacher's mouth, which is the single best thing for the consonants English speakers mangle (looking at you, "zh" and "x").

The catch: clicking the chart is free, but the videos, the PDF download, and the lessons all sit behind an account. Good product, but it wants you in its funnel.

Arch Chinese — totally free, no account. Click a syllable and it shows which of the four tones actually exist for it, because not every syllable has all four. That's a small detail most charts skip.

Downside: the UI is dated, and the audio is a bare recording — nothing to help you shape the sound.

Yabla — free, clean, native audio. Its real strength is the tone-pair drills attached to it, which are arguably more useful than the chart itself once you know the basics.

One thing to know: Yabla lists some syllable combinations that aren't real words. Fine for drilling, misleading if you assume every cell is vocabulary.

Chinese Pronunciation Wiki (AllSet) — a lot of serious learners call this the best chart there is, and I get it. It shows pinyin, IPA, and zhuyin together, with a full-screen mode that makes it feel like a reference manual instead of a web toy.

The tradeoff: it's dense. If you don't already know IPA, there's a wall of notation between you and the sound.

What I built

The Chinese Path pinyin chart is deliberately boring: every Mandarin syllable on one page, click any one and hear all four tones, no account, nothing to download, nothing to install.

I made it because the tools above are all great for studying, but none were great for the thing I actually did most: a thirty-second check between flashcards. I wanted something I could open, click, hear, and close — no login, no video buffering, no notation to decode.

It's not the most feature-rich chart on this list, and it's not trying to be. That's the point.

The short version

Chart	Free	Needs account	Click to hear tones	Standout
Yoyo Chinese	Basic yes	Videos & PDF	Yes	Video demos of tricky sounds
Arch Chinese	Yes	No	Yes	Shows which tones exist
Yabla	Yes	No	Yes	Tone-pair drills
Pronunciation Wiki	Yes	No	Yes	IPA + zhuyin, full-screen
The Chinese Path	Yes	No	Yes	Zero-friction, all four tones

My honest take

If you're brand new and struggling with a specific sound, watch Yoyo's videos — nothing beats seeing a mouth. If you're an IPA person, the Pronunciation Wiki is your chart.

But for daily drilling, pick the thing you'll actually open. For me that's The Chinese Path, because it asks nothing of me and gets out of the way in under a second.

Pronunciation is reps, not tools. Any of these will get you there if you use it every day.

Search engines saw 4 syllables. The page had 199. Here's the bug.

superfire — Sat, 17 Oct 2026 04:47:51 +0000

This article was created with the help of AI and reviewed line by line by the author. Every number below was verified against the running site.

A pinyin chart is a table of roughly 400 syllables. The one I maintain had 199 of them in its data file.

Both numbers were true. Bing agreed with the first one and disagreed with the second.

The bug we shipped

The page rendered its table at runtime. The HTML that a crawler fetched contained an empty table element: a table tag with the class pc-grid and nothing between its tags.

Everything below that point (19 initials, 199 syllables, the four-tone grid) lived inside a JavaScript object called SIMPLE. The code was correct. It just did not exist in the document.

Strip the script tags from the served page and you can read four syllables in its text content. Four.

What the numbers said

Bing Webmaster Tools, three-month window. One page, 101 impressions, 0 clicks, average position 6.

We had assumed this was a copy problem for weeks. Title, description, the promise in the H1, all of it fine. A page sitting at position 6 for its main query with zero clicks out of 101 impressions usually means one thing: the searcher arrives expecting a chart, finds an empty box while the script boots, and leaves before it paints.

The fix

Pre-render the data into the markup at build time. Keep the JavaScript for the click handler, drop it from the rendering job.

The build step walks the same data structure the client script would have walked, and writes each syllable into a table cell as plain HTML. The served document now exposes 374 syllable cells to a crawler, and the page still plays audio on click because the JavaScript never went anywhere.

Same treatment on four more pages in one pass (tone comparison, name generator, stroke order rules, number hand gestures). One rule held throughout: if the content is the point of the page, it cannot live behind a script tag.

The page in question is https://thechinesepath.com/pinyin-chart if you want to check the markup yourself.

The part that bit us

The obvious version of that script was not idempotent. It ran over five pages, the name generator threw on a missing anchor, and the first four files were already on disk. Re-running wrote them again. One data table ended up with five copies of a row where it should have had three.

The rule became: dry-run every page first, and if any single one fails, write nothing. Mark what you injected with a token you own, never with something you count in the output, since counting injected markup to decide whether it was injected cannot work.

One more thing about reporting

The same project reported "zero backlinks" for nineteen days. The tool returned an empty list every time, so zero went into the notes and stayed there.

The backlinks existed the whole time. Three links from a DEV post, all live, none of them nofollow. A different endpoint in the same API (one returning an anchor count per URL instead of a list) showed them immediately.

An empty list is not an empty set. Before writing any number into your notes, check whether your tool is reporting "nothing found" or "nothing searched."

Why this is worth writing down

If you serve a static site and a chunk of your content is generated client-side, run this check today: fetch your own page, strip the script tags, count what is left. If the number is small, you already know why your rankings look flat.

Show DEV: 25 free Chinese learning tools with no backend, and the honest traffic numbers

superfire — Mon, 05 Oct 2026 14:41:37 +0000

What it is
I built this because I couldn't find a free place to practice the boring parts of Chinese. Typing tones. Getting stroke order right. Knowing which measure word goes with which noun.
It's 46 pages now, and 25 of those are tools or resource pages. The pinyin chart is the one people use most. The stroke order page animates each character. There's a transport simulator that runs real airport and subway conversations, plus a hospital one, because those are the two situations where I froze hardest.
All of it is static HTML, CSS, and vanilla JS. No backend, no database, no login page. The audio is the browser's own speech synthesis. I call it that instead of dressing it up as an AI tutor.

How the traffic looks

I'd rather post the real numbers than a win story.

Eight days of Cloudflare data: 93 page views, 71 sessions, 66 of them real humans once I strip the bots. About one visit in nine came from a search engine. Zero backlinks.

That last number is the one that matters and it's the one I can't buy.

Ranking on Google is out of reach from where I build, so I'm working Bing, which is maybe a tenth of the English market. What gets indexed, gets indexed slowly.

One thing did catch me off guard. A ChatGPT share link turned up as a referrer. Somebody pasted a page into a conversation and clicked through. It's nofollow, so it earns no ranking credit, but it means the content is reaching people somehow.

Three weeks in, mobile was at 50% of sessions for a day. Four countries, all phones.

People learn Chinese on the bus, not at a desk.

What's next

The 500-phrase PDF is the only paid thing, at $9.90. Everything else stays free.

If you've run a domain with no backlinks and no traffic, I'd like to hear what moved the needle for you.