First of all, two things.
One, the second we bring in the length of the article into the picture, we are probably taking away the essence of it — the content, which should always be the prime focus. If to get your point across, you need just 2 mins, that works. If you need 10 mins, and you can still maintain the interest of the users till the end — go crazy. Why not!
Writing valuable content and keeping your readers spellbound has got nothing to do with the length of the segment. A badly written piece — no matter how short — will still be bad. Same goes for its counterpart. That is why you can feel the same rush when you read the best authors of the world — whether you are reading a letter they wrote, a novella or an epic. They command your attention — from start to finish.
Two, as far as these two segments of the original story goes, I couldn’t agree more. They are bang-on (other than that one advice to look for stuff to pad the seams with, if the story is shorter 😉)
If you haven’t read the original story, it may be a good idea to do so.
Also, note to Thibaud Herr — Buddy, maybe your approach is right, maybe it isn’t. I really don’t know. I have my doubts, and that’s all that it is. My response got longer than a couple of lines, but I would love to know your views.
WHY THIS APPROACH “MAY NOT” BE RIGHT
I am not sure we would be looking at it the right way if we approached the data in this manner.
I may be wrong, so take it with a pinch of salt. 🤗 🙃
Sure, on the face of it, the graphs make for quite a string of compelling arguments, but I think we would overlook the possibility that this approach is leading us in misinterpreting data? Or reading too much into it — which would be a better way to put it I suppose.
Data misinterpretation? Hashtag-blasphemy!
I could just easily make the argument that you are missing on some crucial aspects:
#1. THE SNOWBALL EFFECT
Momentum does matter. A lot.
The time it takes for your story to go from 0 to 20 and then, maybe 50 recommends?
I am sure there would be a ‘exposure’ event (say, how much of a wider audience does this story need to get exposed to) that would get triggered with each milestone.
And every time it gets exposed to a wider audience, the story is
now reaching out with
PRIOR VALIDATION. We are all susceptible to social norms. And if the triggers are being activated in a short period of time, the validation is even stronger. What does that mean? The chances of that story getting viewed, and even recommended just increased by a significant factor.
#2. THE TIPPING POINT
A distant cousin of the snowball effect. But once this barrier has been breached, there is just no turning back.
Call it anything — the viral loop, crowd frenzy or what not; fact of the matter is for everything a tipping point exists. A threshold on crossing which the traffic just explodes. More views, more reads, and more recommends.
#3. A FOCUS ON THE WRONG METRICS
Now this is where I genuinely let the crazy loose. I did some quick searches to fetch few publicly available Medium stats-page screenshots. Have a look at one of them.
Few points to note:
- Though as a general rule of thumb, it’s said (and is true as well most of the times) that posting your stories on publications gets you better results, as you can see, it is not necessarily true. Content and the audience’s reaction to it trumps all other rules. Every. Single. Time.
- Crafting links underline is a ten min read for what appears to be a simpler topic. Yet the audience response is nothing short of overwhelming. (I haven’t read it, so I can just speculate on the simplicity of the topic, but obviously there is something in the content there to garner such traction)
Anyway. Circling back to the wrong metrics, let’s try to see the ratios here. Medium already tells you the read vs view ratio. Though I’m not really sure of how it works. Does it count the whole article being read, the story just being opened in full view or a certain minimum percentage content read.
I’ll additionally try to see the read vs recommend ratio.
Irrespective of the trigger for counting a view, if a person has viewed it, a useful story would typically get a heart. (Sure, just because the story was found useful by the reader, the reader may not always be in a generous mood for a heart, but still we can keep the general pattern as such)
So this metric becomes important IMHO.
The one with the highest total number of recommends — Crafting links — has a 3% recommend ratio. Which means that only 3 out of every 100 people reading the story:
- stuck till the end
- found it useful
- found it insightful
- could not move on without feeling the need that this was a story worthy of being read by others
Couple of points to be noted here:
- The stats page screenshot was fetched from a google image search result.
- It would be dated. I searched for some of these stories, and their current traction numbers are quite higher than what it shows in the screengrab here
- I have used a simple x1000 to convert Medium’s 4.9K data to total number of views/reads/recommends. The actual number would be higher.
Now, let me show you a different picture.
This is a screengrab from my own stats page. I have a couple of other stories as well which have a recommend ratio of 33% or higher — as the one shown in this picture. Does this mean that story is good? No, I don’t think so.
My typical stories have a recommend ratio between 12% — 20%. Does that mean that people find my stories insightful? Again, I don’t think so.
So what the hell is going on. I’m glad you asked, and I will answer that in a min.
To be honest, I really can’t say for sure if using recommend ratio is the best route here or not. The point I was trying to make is — you can manipulate representation of data very easily to prove a point.
NOT DISCOVERED != NOT USEFUL
Just because it has not been discovered yet doesn’t mean it is not a content that is useful, and would be appreciated by many once discovered. Biggest example, Franz Kafka.
I think I should make it clear before moving on — No, I don’t think the reason my stories have low traction is because they haven’t been discovered yet. I am not that vain. :-p
Kafka is widely regarded as one of the most gifted literary minds of the 20th century. So much so that there is an actual term in the English language named after him — Kafkaesque.
Why do I think it’s interesting enough to bring it up here? Because no one hardly acknowledged his presence (let alone how gifted he was) during the time he was alive. He lived, and died, in relative anonymity and obscurity. One of his best works — The Metamorphis (Die Verwandlun) — has been treated as a gospel by writers and critics for more than half a century, and yet it never received any acclamation during his lifetime, despite being published in literary magazines.
I was working with a leading e-commerce business. Well funded, good market share share and all. The first problem I was asked to solve was the huge contribution of stagnant inventory in their system. A well funded business, dead inventory. What is the first thing that comes to mind? Yes. DISCOUNTS!!!
Now, A. As a professional, I hate giving discounts for the sake of giving discounts. And B. They had already tried that a number of times before they even brought me onboard. It was a temporary fix at best, and not a very effective one at that — the results weren’t what they were expecting.
So, how do we approach this problem? Lets look at data! We fetched a lot of raw data from the system — primarily to evaluate if a correlation exists between how much a product is displayed and how many units of that product gets sold in a particular time. While trying to look at that, we accidentally noticed what we would later call ‘min. visibility’, and later ‘the ideal exposure matrix’. There were products on the system that consumers were simply unable to discover. They weren’t being displayed effectively. And they can’t buy what they can’t see, can they? So, that was the angle we approached this exercise from, and the results were great!
Discovery matters. Almost as much as the quality of your product!
ALRIGHT. SO WHAT’S THE POINT OF ALL THIS RAMBLING OF MINE?
Fuck if I knew!
- I don’t know what should be the ideal length of a story
- I don’t know what is the best time to post a story — but sure, assuming an x% of traffic is coming from North America, it makes sense to stick to that schedule. But I don’t have enough data to make that recommendation
- I don’t know how to get noticed, how to build a following, how to become a top writer.
All I do know, for sure, is that content rules!
Relevant. Contextual. Engaging. Entertaining. Informative.
Just came across this story earlier today (accidental bump in. 100%). Possibly posted a couple of hours before I noticed it first. I had a feeling it may help illustrate a point, so I took a screengrab.
Why am I showing these screengrabs?
- The story was posted at a time when half of Medium’s audience is getting ready to go sleep (I’m not sure how accurate this statement is; I am basing it on what I read once — the best time to post a story on medium is between 6:30AM — 9:30PM central time)
- The story is FUCKING long. 24 mins? Are you kidding me!
- And yet, it managed a runrate of almost 2 recommends for every hour.
Content. That’s the only game in town — just like it has
ALWAYS been. Medium is not 100% new. There have been good, clean blogging platforms in the past as well. What Medium is, is a new mode of delivery as well as discovery. And that is why it works; and it does work great! When it comes to discovery and/or delivery, Ev Williams and his team has been working on something that is amazing. But whether you had a number of unfinished, unpublished manuscripts like Kafka had, or a hyperactive Medium page, content reigns supreme.
Length of the story? Really doesn’t matter.
That’s it for today! See you tomorrow…
As I finished with the content of my response, it was a 8 min read. Really! You can’t make this up.
But I am yet to add my Upscribe code and my contact info, and then there is this ‘irony’ segment. So, I think it would be 9–10 mins at the end.
WHY DID I WRITE THIS STORY?
I have read arguments for 6 mins, 7 mins, and now 8 mins; but how can that be? How can we assign the ideal length to a story? Wouldn’t it always vary with context? Subject matter? Audience type? Platform?
Okay, I have to give you platform here, because wherever I have read these numbers, the platform in question has been Medium, but what about the other critical factors?