Content Analyzer

A content analyzer fetches a page and reports how well its on-page elements are set up for search and for readers. It reads the title tag, the meta description, the heading structure, the images and their alt text, the internal and external links, and the body copy, then measures each against a published threshold and returns a score. It is the same pass an SEO would make manually on a page, done in a couple of seconds and applied consistently.

The seven checks are the ones with a defensible number behind them. A title between 30 and 60 characters is what Google renders before truncating; a meta description between 120 and 160 is the same constraint applied to the snippet. Exactly one h1 is the HTML outline rule rather than an SEO superstition. Six hundred words is the point below which a page usually cannot cover a topic well enough to satisfy the query. Alt text is an accessibility requirement first and an image-search input second; the check passes when no image lacks alt text, warns when up to a fifth do, and fails beyond that. Three internal links is a floor, not a target, and it exists because orphaned pages are the single most common structural problem on a site.

Readability is scored with Flesch Reading Ease, which weighs average sentence length against average syllables per word. Sixty to seventy is plain English; below thirty is the register of a legal contract. The readability check passes at 50 or above, warns from 30 to 50 and fails below 30, so a page does not have to reach plain English to pass. The syllable count is a heuristic and it is wrong on individual words, but across a page of prose the aggregate is stable and useful. Alongside the score you get the ten most frequent meaningful words, with stopwords removed, which is the fastest way to see whether a page is actually about what you intended it to be about.

The limits are worth stating plainly. It reads the HTML the server returns, so a page that renders its content with JavaScript will report almost no words and fail several checks that it would pass in a browser. The word count, readability score and keyword list are taken from all the visible text in the page body, so navigation, sidebars and footer text count along with the article, and a page with heavy site chrome reads longer than its main copy. It follows up to five redirects and checks every hop before requesting it, so a URL whose host resolves to a private, loopback, link-local or other internal address is refused without being fetched, as is any port other than 80 or 443. Pages that declare a size over three megabytes are rejected, and longer responses are cut off at that size. And a high score is not a ranking: these checks measure whether the page is set up competently, not whether it deserves to outrank anything, which depends on the content and the competition rather than on the markup.