See whether a crawler reads your h1 as whole words or as loose letters.
Fragmented
What a person sees
Build something worth reading
What a tag stripper gets
B u i l d s o m e t h i n g w o r t h r e a d i n g
Up to 100,000 characters. Pasted code is never run.
<h1>Build something worth reading</h1>URL checks fetch the public page on the server. The HTML playground runs entirely in this browser; pasted source is never uploaded.
How it works, and what it can't see
Both readings are models, not claims about any named crawler. The first joins the heading's text the way a browser lays it out. The second also breaks at every element a plain HTML-to-text pass would separate: an inline element given a block display in its own style attribute, and a letter in an element of its own when three or more build the heading.
Pasted HTML is read as source text. Nothing in it is run and no page is rendered, so a heading that a script adds later is not there to read.
Why this is worth ten seconds
Split a headline into one element per letter and a browser still paints the word. Plenty of the machinery that reads the web does not run a browser: link unfurlers, feed readers, archivers, and the fetchers behind AI answer engines. A good number of those strip the tags, normalise the whitespace, and hand the result to something else. That turns your best string into confetti, and nobody sends you a report about it.
I found this on my own site, which is the only reason I trust it enough to write a tool about it. The homepage name animated one character at a time and extracted as loose letters. It came out of building a CRT that behaves like a CRT, which is where the same warning sits at the end.
Can't see
- Your stylesheet. It reads the served HTML and the style attributes in it, so a class that sets display:inline-block is invisible to it. One element per character is the signal that survives that.
- Anything JavaScript renders after load. If the heading arrives from a script, the served HTML has no h1 and that is what it reports.
- Pages behind a login. It fetches as a stranger with no cookies, so whatever a visitor has to sign in for is out of reach.