Write Readable and Searchable, With or Without AI

When we talk about text today, the focus often lands on whether something feels AI-generated or not. But that’s a pretty flimsy way of looking at things. There have long been established measures and approaches that still hold up. Readability is one of them. Is the text enjoyable, meaningful, and easy to absorb? And when it comes to those search engines – well, that readability takes you pretty far.

In literature and language studies, for those who like to get technical, there are various metrics for readability. One old but still relevant one is the Readability Index (LIX). Sure, it’s been mocked now and then, but as part of the bigger picture, it’s far from useless. It looks at average sentence length and the frequency of long words. Basic, maybe, but meaningful when used in the right context.

There’s also word variation indexes, and measures of text cohesion. Especially looking at transitional words or logical connectors like “therefore,” “instead,” “on the other hand,” “meanwhile,” “despite that,” “rather than.”

And then there’s the more visual layer, breaking things down with bullet points and numbered lists. This can help or hurt depending on how it's used. But it’s clearly important for searchability as it can be for human perception.

Naturally, there are different types of texts – fiction, academic, instructional – and each should be judged differently. But overall readability is still the most interesting lens. Why? Because the boom of “machines writing for other machines” probably won’t last long.

This thinking is deeply embedded in how we at KontentPlus analyze and create content. We look at the bigger picture, not just whether today’s search engine algorithms happen to pick up one keyword over another — even if SEO (Search Engine Optimization) and LLMEO (Large Language Model Engine Optimization) matter too.

And no, we don’t settle for vague mantras like “good content is good content.” We measure with a wide range of metrics to assess the readability and impact of the content we deliver. Whether it was created with AI support or not? Not really the point.

Quick Example – Readability Snapshot of This Text:

Here’s a small sample of the metrics used at KontentPlus:

• LIX (Readability Index): 35.8
  Reference: 30–40 = Easy, 40–50 = Medium, 50–60 = Technical, >60 = Difficult
  → Easy-to-read, near-perfect for a broad audience.

• Average sentence length: 15.2 words
  Recommended: 12–22 words
  → Well within the ideal range, contributing to a smooth reading experience.

• Percentage of long words (>6 characters): 20.6%
  Norm: 20–25%
  → Right on the threshold — a good balance between simplicity and informativeness.

• Lexical density (content words): 53.7%
  Norm: 40–55%
  → On the higher side, but still within range for easy comprehension.

• TTR (Type-Token Ratio / word variation): 0.55
  Healthy range: 0.4–0.6
  → Good vocabulary variation, avoids repetitive fatigue.

• LLMEO Score: 80/100
  Goal: ≥70 = Good, ≥85 = Excellent
  → Very strong structure; could be further improved with additional subheadings and bullet lists.