How to Estimate Reading Time Before You Publish

You have finished a draft and want to put a reading-time label beside its title. The difficult part is not division. It is making the label describe the page a visitor will actually read. Copying a whole webpage into a counter can include navigation and related articles; counting only the introduction leaves most of the work out.

This guide follows a hypothetical 1,000-word article through counting, labeling and revision. It is an editorial method, not a measurement of your audience. Use the linked Reading Time Calculator to count text and apply a pace; use the comparison article when choosing between 200 and 250 WPM.

Make a plain-text copy of the main article. Include the title, meaningful headings, prose, captions and list text readers need to understand it. Exclude the site menu, footer, cookie controls and unrelated recommendations. Decide once whether the title is included and apply that choice consistently across your publication.

Consider an illustrative export containing 1,000 article words, 80 words of navigation and 120 words of recommendations. Counting all 1,200 words produces 6 minutes at 200 WPM; counting the article alone gives 5. Removing the extra minute is appropriate because you removed unrelated text, not because you chose a faster pace to make the article look shorter.

Treat 200 WPM as one transparent house assumption, not the correct speed for every reader. Compare 250 WPM if you want to see the effect of a faster assumption. A lower pace may be appropriate for a planning estimate with unfamiliar terminology, but lowering the number does not turn the formula into a measured assessment.

Keep the pace consistent for comparable articles. If two editors use different defaults, their labels are not directly comparable even when both calculations are correct. Record the reason for an exception. For a speech or video script, use a speaking-time workflow and rehearsal rather than borrowing a reading label.

Raw minutes = counted words ÷ selected WPM. The tool displays the next whole minute for a non-empty input: 1,150 ÷ 200 = 5.75, displayed as 6 minutes. Empty text returns 0. The upward rounding is a display rule, not evidence that the reader needs exactly that long.

Paste the selected body text, check the reported word count, choose the pace and calculate. If the word count differs from your editor, compare token boundaries before changing the pace. Hyphenated words, punctuation and writing systems can be counted differently. The related word-count guide explains why counters disagree.

A tutorial with a five-minute prose estimate might also ask readers to install software, inspect a chart or complete an exercise. Those activities are not represented by words divided by WPM. Do not silently roll them into a precise-looking reading number.

A clearer label is ‘Estimated reading time: 5 minutes; exercises take additional time.’ If you actually time a complete walkthrough, record the device, steps and conditions before publishing that as a separate observation. Do not add an arbitrary number of seconds per image and call it tested.

Use wording such as ‘Estimated reading time: 5 minutes.’ Keep the label near the title or introduction so it helps someone decide whether to start now. If you publish your method, explain the counting scope and pace in plain language. ‘Calculated at 200 words per minute; activities excluded’ is more informative than a bare number.

Before publishing the final version, run the count again. In the handoff example, adding 150 words moves the rounded label from 5 to 6 minutes. A minor punctuation edit may not change the label, but changing the method without explanation makes the archive inconsistent.

Record the final word count, selected pace, raw result, rounded label and any excluded activity. This takes five short fields, and another editor can reproduce the decision without needing your browser session. Keep confidential drafts within your organization's permitted tools and avoid including real private text in an example or shared review record.

The final check is human-readable: does the label describe what the page asks the reader to do? If the answer is no, explain the extra work or remove the label. A formula is useful only when its scope matches the promise.