Adding a transcript word count to your podcast About page

Sunday, 14 June 2026

If you have told Podcaster where to find your podcast audio files, it can provide you with some fun facts about your podcast that you can include in your podcast templates.

A good place to include this data is on your About page. Here’s an example from the 500 Year Diary website’s about page template.

## Stats

There are {{ podcastData.numberOfEpisodes }} episodes of _500 Year Diary_, 
weighing in at {{ podcastData.totalSize | readableSize: 2 }} 
and lasting {{ podcastData.totalDuration | readableDuration: 'long' | replace_last: ', ', ' and ' }}.

Here’s what the output looks like:

As I was adding transcripts to my podcast websites, it occurred to me that there was another fun fact I could include on the about page: the total number of words spoken on the podcast.

So here’s how I did that.

Calculating a word count for the transcripts

First, a reminder of the VTT format of the transcript files. Here’s an excerpt from the transcript of Season 4, Episode 5 of 500 Year Diary.

WEBVTT

NOTE
This transcript was created on 2026-06-07 at 09:18:04

1
00:00:33.359 --> 00:00:37.679
Hello, dear listener, and welcome back to 500 year diary.

2
00:00:37.740 --> 00:00:46.140
The only Doctor Who podcast that's just thought of an excellent way you can get out of buying Christmas presents for your nieces and nephews this year.

3
00:00:46.200 --> 00:00:47.880
I'm Nathan.

4
00:00:47.939 --> 00:00:48.659
I'm Adam.

5
00:00:48.719 --> 00:00:50.219
And I'm Todd.

6
00:00:50.280 --> 00:00:53.820
It's Friday, the 10th of July, 2009.

Here’s a summary of the format:

Each VTT file consists of blocks of text separated by blank lines. The first of these blocks is a WEBVTT header, which tells the browser that this is a WebVTT file. After that, there are two kinds of block — note blocks and transcript blocks. Note blocks are just plain text, but transcript blocks consist of:

• an optional line number
• a required timestamp range consisting of two timestamps separated by an arrow (-->)
• some lines containing the actual transcript content

Here’s a JavaScript function that takes VTT content, extracts the actual transcript content, counts all the words, and returns the result:

function countWords (vttContent) {
  const text = vttContent
    .trim()
    .split(/\n\n+/)
    .filter(block => block.includes('-->'))
    .map(block => {
      const lines = block.split('\n')
      const timeIndex = lines.findIndex(l => l.includes('-->'))
      return lines.slice(timeIndex + 1).join(' ')
    })
    .join(' ')

  return text.split(/\s+/).filter(word => word.length > 0).length
}

If we know the path to the transcript directory, we can use this function to count the words in all of that directory’s VTT files.

import { readdirSync, readFileSync } from 'node:fs'

const TRANSCRIPTS_DIR = './src/transcripts'

export default readdirSync(TRANSCRIPTS_DIR)
  .filter(f => f.endsWith('.vtt'))
  .reduce((total, file) => total + countWords(readFileSync(`${TRANSCRIPTS_DIR}/${file}`, 'utf8')), 0)

To make that available to our templates, we can include it in a global data file in the _data directory. Here it is:

_data/transcriptWordCount.js

import { readdirSync, readFileSync } from 'node:fs'

const TRANSCRIPTS_DIR = './src/transcripts'

function countWords (vttContent) {
  const text = vttContent
    .trim()
    .split(/\n\n+/)
    .filter(block => block.includes('-->'))
    .map(block => {
      const lines = block.split('\n')
      const timeIndex = lines.findIndex(l => l.includes('-->'))
      return lines.slice(timeIndex + 1).join(' ')
    })
    .join(' ')

  return text.split(/\s+/).filter(word => word.length > 0).length
}

export default readdirSync(TRANSCRIPTS_DIR)
  .filter(f => f.endsWith('.vtt'))
  .reduce((total, file) => total + countWords(readFileSync(`${TRANSCRIPTS_DIR}/${file}`, 'utf8')), 0)

And so now we can use transcriptWordCount in our templates. Like this:

## Stats

There are {{ podcastData.numberOfEpisodes }} episodes of _500 Year Diary_, 
weighing in at {{ podcastData.totalSize | readableSize: 2 }} 
and lasting {{ podcastData.totalDuration | readableDuration: 'long' | replace_last: ', ', ' and ' }}.
The transcripts of all the episodes add up to {{ transcriptWordCount }} words.

And here’s the output:

And we’re done. Silly, but fun. As I’m writing this, Flight Through Entirety has a transcript word count of 3,033,973, and on Untitled Star Trek Project it’s 2,327,981. Am I wasting my life? Or am I hanging out with my friends chatting about my favourite TV shows? It’s impossible to tell.

• ← Previous
  Creating interactive podcast transcripts