Edited by my beloved friend K.L
On September 2022, as usual, I released two articles on my blog. These are two parts of one interview.
Although I cherish these articles and I am enormously grateful for the conversation that I had with my guest, these two articles were the longest to write and the most energy-consuming since the start of my blog. And it was all because of my imprudence.
I started the interview section on the blog believing that… it would be quick and easy thing to do, almost like a small, pleasant recreation from writing my usual articles.
I was imagining meeting a person who inspires me and have a talk, and then I would just have to edit the interview a bit to get the content ready; there would be no need for research, no need for an original though-provoking perspective, no need to turn into a far-reached memory to extract the essential without offending the person who was involved…
I met Lan, my guest, in the spring (of 2022?), thinking of featuring her as a summer character for my blog. However, I could not finish the article until the fall. And it took me 45 hours to get this 2-parts interview ready, which is the equivalence of 22 hours and 30 minutes per article!!!
How did a piece of content that seems “easy” become the most complicated content to produce? I will explain everything in this article. And most importantly, I will give you some tips on how not to make the same mistakes that I did!
The raw pearl
Lan was the very first person I interviewed.
Since I wanted my guest to be relaxed so I could better unlock the conversation, I set no limits, neither on the subject nor on the time. I chose to record our conversation instead of taking notes, in order to listen to my guest’s stories in full awareness of the context in which they were told later on. Our interview went on for three hours in an emotional and caring way.
The moment I stopped the audio recording, I felt like I was holding a raw pearl in my hands.
It was February 2022. I planned to publish the article in July. Naively and recklessly, I thought, “I have plenty of time left!”
I put the recording files away, being confident that I could start processing them in early July.
In the meantime, in April 2022, I published another two-parts interview for my spring guest, Morgane. I applied exactly the same technique: free talk with voice recording. The discussion also lasted three hours.
Writing this spring interview articles made me realize that, quite the opposite of what my preconceived beliefs had been, polishing a raw pearl takes just as much time as writing a traditional article. Interviews also require more responsibility because the guests’ image is now involved.
For that reason, I decided to work on Lan’s interview right right when I was done with the article for my first interview with Morgane.
From the first time I listened to the recordings, I encountered a huge problem…
Since the exchange was carried on in all spontaneity, there was no guideline. Also, I had no experience interviewing people, so I did not know how to canalize the discussion. My pearl was very raw with a lot of information that were mixed up and many subjects that went back and forth between them.
It was impossible to edit as is. It had to be restructured.
We were in May. Three months after the interview, my memory was no longer fresh, and all the ideas for the hook and structure that I once had in mind at the time of the interview were gone.
Listening to the three hours of recording again made me even more panicked. I am basically not an auditive person. Every time I listen to an audio book, I feel lost. I lose track, I miss information, I cannot even detect transitions. You can imagine my struggle with the three hours of audio recordings without structure…
I needed a visualization! I needed a written version!
When “easy” content becomes the most complicated content to produce
The most obvious solution was to transcribe all the audio files into text.
However, there was a new problem: my guest is bilingual in French and Vietnamese, and so I am. If you have ever witnessed a conversation between two bilinguals, you will know that words and expressions from the two different languages are often mixed in the same sentence, like a cocktail. Me and my guest even added English expressions to the discussion.
No automatic transcription software can handle this kind of file.
So I had to transcribe using the traditional technique. That is, to listen, and rewrite what I heard.
In the past, transcribing a 6-minute YouTube video in a single language took me 1 hour. So in all reasonable logics, if it was 3 hours of recording, I would have to allow myself somewhere near 30 hours…
I could not entrust the transcription to another person either, due to confidential reason.
So I changed my strategy. I needed another way to find the structure of the article, then I would have to transcribe only the passages that were essential to the structure.
And in order to better figure out a proper structure for my articles, I needed to pinpoint passages that could be potentially essential!
Preparing the construction site
I listened to all the recordings again, but this time I wrote down the topics mentioned, the information, the anecdotes, along with the timecode (of the audio) corresponding to each item.
At this stage, I could already detect and ignore off-topic passages, information that were too personal or interruptions that brought no added-values. And I made use of my train trips in May and June to get this step done.
After 6 hours of work, I got 28 A5 pages of bulleted list, but still no obvious structure.
All the items were linked together. But in my head, they were still like headphones strings that get tangled in a pocket. If you just take one end and pull on it mindlessly or roughly, you will make a knot, and everything will be a blocked mess. A more careful disentangle work is still necessary for me to get things straight.
The visual solution
I am a visual person. I needed something more visual than just bunch of words sticking together no matter how structured they seem to be! I needed a color code!
I pulled out the post-it notes to assign a color to each topic that we discussed. I then categorized each line on my 28 pages of bulleted lists by theme, sticking the post-it notes with the color that corresponds to the theme next to it.
After 1 hour, all the information has been classified into eight themes (along with it, two thirds of my stock of post-it notes).
A new question arised: In which order should these eight themes be presented to my readers? All the elements are linked, and for each decision that my guest took in her creative path, it would take several information to understand fully in its context. Therefore, I need something else, something that could be even more visual than just colorful post-it notes to bring the real big picture to my eyes before I could attempt any writing.
So, I built a mind map to trace the links between the different themes as well as between elements in each theme (cause – consequence, if – then, why – because…).
The structure showed up naturally after 1 hour of mind mapping.
Transcription: The mountain to climb
It was August when I started the transcription. I decided to translate right away the passages where we spoke in French or in English into Vietnamese in writing, and followed the structure defined previously by going directly to the timecode corresponding to each element.
Even if this step remained quite repetitive and uncreative, getting over this mountain helped me to find the fastest technique to transcribe:
- Slow level: Listen, then rewrite what I heard
- Medium level: Listen, then rewrite, in only one language, out loud, letting the Notion application do the automatic transcription of my voice.
- Fast level: Same technique as level 2, but increase the speed of the audio file to x1.5
My transcript was ready after 10 hours of work, instead of 30 hours.
What was better is this structured version only needed spelling correction and formatting to look like a blog post.
However, a new problem jumped at me: My transcript contained… 10,000 words! And that’s definitely way too long for what I had in my mind for my blog posts.
Cut, cut, and cut again
«Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away. »Antoine de Saint-Exupéry
It was laborious to cut out passages that were original or contained a lot of emotions from my guest but did not fit into the subject of “creativity.” I felt guilty with each deletion. But the moment the final version appeared, the confidence filled me up. The final version came out “perfect”. And in my heart, I felt grateful. Nothing is wasted. The passages that were cut will remain forever in the memory of that spring afternoon, when I met such an inspiring person as Lan is. These emotions, this atmosphere, this confidence, all of that were only for me, at this moment of crossing of two creative paths.
It took me 8 hours to edit and correct the mistakes from the automatic transcription.
By that time, it was September, and I still had to prepare the French and English versions before I could make this interview available on my blog.
The French translation was particularly difficult for me, simply because it is the opposite of what I used to do on the blog (I used to write in French first, then translate into English and Vietnamese).
I once again dictated my translation out loud and used Notion to provide the written version. But this time, it did not save me much time, because the spelling and grammar correction was laborious, the automatic transcription did not have a proper treatment of French grammar, not to mention that sometimes words can be pronounced the same but are not written the same way.
The translation from Vietnamese to French took me 10 hours, including 5 hours and 30 minutes just to correct the spelling mistakes.
The translation from French to English was much faster. I only spent 3 hours on it.
Finding and editing the photos to illustrate the article took me another 3 hours.
I spent 45 hours for 2 blog posts….
This experience taught me a lifelong lesson
An idea that seems simple can hide complexities beyond my capabilities!
What can I do to avoid falling into the same situation?
Before the interview:
- Define the subject: Nothing prevents me from contacting the person again to talk about other subjects in the future;
- Prepare and structure the questions, without being rigid. It’s always better if the discussion leads me in a new direction, but a guideline, even a loosened one, would be helpful for my future tasks.
During the interview:
- Block an hour and a half maximum for each interview. Nothing prevents me from continuing to talk freely with the guest after the interview;
- Think “podcast”, not “blog post”: Imagine a listener who is listening to my audio recording to keep a thread throughout the discussion, and rephrase what was in foreigner language or any sorts of jargons directly during the discussion;
- Above all, no juggling between several languages. My guest has the right to speak in his or her own way, but I am the “host of the podcast”, and so it is up to me to manage my language.
After the interview:
- Establish the structure of the article within the week following the interview, when the memory is still fresh, and the feelings are still intact.
I spent 45 hours on 2 blog posts for the month of September 2022, but I got another big-time win: Since I wrote both posts at the same time, I finished them both on September 5th. So, the September 15th article was actually planned ahead of time.
And this is the very first time that I was able to use this planning feature, in my 10 months of blogging. J
Tu Ha An