A couple of weeks ago Medium 1.0: A new storytelling experience was announced. While I did enjoy the subtle design changes, the main changes made for a worse writing experience. This regression is due to Medium losing sight of its value proposition. As an amateur writer, Medium solved two problems: (i) the solitary writing journey; and (ii) the post-writing limbo. I am not a particularly good writer, hence the writing process is an arduous one. Once finished with this laborious process, I must embark on the equally unpleasant task of distributing my writing to make it worth the trouble.
Up until now, great writers wrote and others distributed their work. The rare successful self-publishers tend to produce great content and have great distribution channels to promote their work. However, most of us are either good writers, or good marketers but not both.
Medium is not a publishing tool. It’s a collaborative content creation platform.
Medium solved my first pain (the solitary and difficult writing) by creating a fluid writing experience, allowing others to suggest changes, as well as leave notes to improve my ideas. Medium initially solved my second pain (post publishing limbo) by distributing posts through collections where anyone who follows a collection would automatically see my writing as soon as I published in a collection they subscribe to.
This experience made Medium much more than a publishing tool. It was on pace to become a scalable collaborative content-creation platform. This difference might seem like semantics, but it’s actually critical. If you aspire to be a publishing platform, then the ceiling is pretty low and has been set by Blogger & Co. But being the first scalable community idea-generation and distribution platform, has the potential to transform how we read, write, and create content for the masses.
Medium has a choice to make: be the next Blogger or become the first community idea generation and distribution platform
The main problem with Medium 1.0 is the changes they made to collections and how you submit your work to them. Currently, sharing stories in collections is an extremely tedious task. To you publish a story in multiple collections, you have to go an unnecessary cumbersome process where you need to: go to the collections page, search for collections using a keyword, click a relevant collection, visit its page, click to apply for membership, select a single story you wish to submit, and click OK to confirm. You have to repeat this process for every collection you want to submit to.
This process is like writing a tweet and then manually selecting who would see it, except you have no good way to browse Twitter accounts or to tweet multiple people at the same time.
The current collection submission mechanism is like manually selecting the users you want to see your Tweet
How to make Medium better?
So how to make Medium a scalable community content creation platform where users are enticed to write great pieces because of the beautiful writing experience and the effortless distribution process? Here are some suggestions:
Forget features — focus on value proposition
The majority of feature requests are incremental, instead focus on the transformational aspect of your value proposition: you will become a better writer because of our beautiful interface, our distribution, and idea co-creation.
This is a powerful value proposition as it could mean that the next JK Rowling won’t be a single person writing alone in a cottage, but a dedicated community that will co-create amazing ideas by building on each other’s creativity. What’s really exciting about this idea is that it will not only take brilliant co-authors, it will also require great readers to provide insightful comments, suggest changes, and provide support. Medium can be the first writing experience where users seamlessly provide feedback to shape the product.
The power of Medium is that it has the potential to create better written stories, with better ideas, and better supporting materials than any single author could create on her own.
So what does writing on a community idea creation and distribution platform look like? Take this essay for instance, it contains neither the best writing nor the best ideas— I simply shared with the broader community a vision for this platform. Now (if Medium solves its distribution problems) others would read this story, suggest edits, augment the ideas within, and the combination of ideas could lead to a novel outlook. Medium could be the GitHub of writing. Ultimately, we would have a better written story, with better ideas, and better supporting materials compared to me writing alone — and that’s the power of Medium.
Collections are a powerful tool to promote your users’ work but they need significant changes. I analyzed 1,000 Medium collections and here are some suggestions based on the data.
1- Submissions: The quickest fix is to overhaul the submission pipeline. Instead of the laborious process outline earlier, just have the interactive search box in the tab allow writers to search for collections and click to submit. This is still problematic as the search feature is very finicky and novice contributors might not know which keywords are best.
1a- Fixing search: If the submission process remains the same, then the search feature must be improved. Currently, search is very finicky. I sometimes type the exact collection name I am looking for and it would return “no results”. Furthermore, it’s unclear how the search algorithm is matching queries. Some results have no apparent relevance to the query a user types. I more robust “search engine” would return results based on some relevance between the keyword, the collection name, the posts within the collection, and the popularity of the collection. It goes without saying, most people are searching for popular and relevant collections to submit to.
1b- Saving submission workflow: Let’s assume the submission workflow remains. Then one helpful function is to save the collections I submit to regularly, or those that generated a lot of reads and recommends and show that subset of collections immediately when I click “Submit to Collection”
2- Multiple “editors”: I find the term editors for people who manage collections a bit misleading. Maybe they should be re-branded as “curators”. As the data shows, the collections with the most followers tend to have the most stories thus an editor must review a lot of submissions. This means two things from an editor’s perspective: First, if they are selective, they are likely flooded with submissions and response times will be very long. Second, if they accept every submission, then the content quality is going to worsen and reader engagement will likely go down. This is why editors are a critical piece of the puzzle. Without quality editors, the Medium experience will suffer. So each collection should have multiple editors rather then just its creator, some of editors could be popular writers in that collection.
3- Editor quality: In addition to the collection information currently provided to writers such as the number of followers. An editor quality score that is a function of response time, follower engagement with the content (s)he curates, and how engaged the editor/curator is within the community should also be listed. This way, editors are held at a higher standard given their important role in the ecosystem and writers know the kind of editor they are dealing with when submitting a story.
4- Decision rationale: When making a decision, each editor should provide a rationale for the acceptance/rejection of story. All co-editors should be able to see this decision and score that decision. For example, a submission might be rejected because it lacked “relevance” to the collection. Other editors can see this decision and should “vote” up or down that decision. This will ensure that people are actually including/excluding stories based on merit and not favoritism or grudges. This is a huge improvement over traditional newspapers where you never know why your Op-Ed story was rejected. People are more likely to contribute to a transparent platform.
5- Auto-submission: One of the reasons Pinterest became so popular was that, for every new user that signed up, they were automatically signed up to follow a few interesting accounts. This caused some users to have millions of followers overnight, but at the same time it increased overall engagement. Given Medium’s data, it’s easy to build a probabilistic topic model that assigns each writing to multiple topics based on its content. A story could then be automatically submitted to collections that have high engagement with the topics associated with the post. It’s a win-win: the writer gets exposure and the reader gets relevant content. Alternatively, you could suggest “similar collections” based on the collection I am submitting to and other collections where there are a lot of cross-posting or collections that share many followers.
6-Exploring collections: As of now, it is almost impossible for me to discover new collections to follow or submit to. In addition to the search box on the collections page, there needs to be a directory of collections — some of it curated and the rest organized by topic or alphabetically (similar to an App Store).
Growing ideas and collaborative ownership
Ideas are never born in a vacuum. We are all influenced by the ideas of others, who in turn had their own external inspiration. Therefore Medium should not try to encourage the delusion that we own our ideas. When we write, we hope to create a better world, but to create better ideas we need to share them, grow them, and watch them evolve.
People or editors can take an idea and improve it by posting notes and suggesting changes. The more popular suggestions could be automatically inserted into the story. Similarly, additional materials (videos, graphics, etc.) can be suggested and contributed to the writing. Every contributor becomes a co-author or co-creator and the winners are the readers who are experiencing a constant evolving and increasingly sophisticated idea. Co-owners can then submit the story elsewhere or invite others to contribute.
This concept of co-ownership and idea co-creation is a powerful one that would set Medium apart from other platforms.
I noticed that collections have a “paywall” flag in their HTML. Paywalls are tricky, but I prefer the story-level paywall, where author(s) charge for a story. This could transform online publishing, especially if you can have different templates for different content (books, articles, comics, etc.) For example, Walter Isaacson is co-editing a chapter of his new book on Medium. One question that comes up is that of compensation for editors. My suggestion for co-owners would solve this problem: the most popular edits and suggestions that get added to the text would become co-authors and will get some percentage of the revenue. But Medium can be more than an “editing” platform (at least more than what Isaacson is using it for), with some design changes one could publish their book on Medium and charge for the content.
Obviously, the story-level paywall will be exciting for journalists who won’t need newspapers and magazines to make money. This will only be possible if Medium excels at the frictionless distribution I am advocating, since it will allow writers to focus exclusively on writing, not marketing.
If these value proposition suggestions are not in line with Medium’s vision, then I have a few vanity feature requests:
One improvement to the stats page would the ability to get information from the bar chart into table format. In the stats page, I should be able to click on the bar graph for the views, reads, or recommends on any given day and it would highlight the posts that contributed to that daily quantity (i.e. bar height). Right now, when I click on the bar graph, as shown above, all I get is this bubble that tells me how many reads I got that day, it doesn’t tell me which posts contributed to the daily read count.
Advanced engagement stats
In the old days, books were published with the hope they might be read but with no data back up that hope. Today, we have an unprecedented chance to have an intimate look into how readers interact with our writing. I enjoy the “# of reads” statistic. I assume that a visit is counted as a read when a visitor spends at least the same time as the expected reading time and/or reaches the end of the page. What would be useful is to develop stats that quantify how readable my content was: (i) where did the most people exit my page?; (ii) are there some words/sentences that many people were highlighting? (iii) were there sections that people were spending a lot more time on? Such stats would allow a writer to improve certain sections or identify segments that people are enjoying.
Do not allow people to host Medium pages on custom domains
This is a feature “un-request”. I know many people will want to have their Medium page to point to a custom domain, like WordPress does. But just like Twitter, you want to create a vehicle, a protocol, and an environment where ideas can seamlessly originate, be shared, and improved. Medium’s differentiating factor is it’s community experience and allowing custom domains reverts it back to the traditional publishing platforms that are readily available.
Cover photo repositioning
I love the cover photos. One problem is that we can’t reposition the photo. Please allow us to reposition a photo in cover/landscape mode as needed.
If you found value in this article, I would be grateful if you hit the recommend or Tweet button below.
Thank you to: Michael Brewster, Anand Rao, Raj Menon, Eroteme Thinks, Morten Wang, Jackson, Andreas Mitschke, Scott Wilkinson, Philippe Kuhn, Vicente Plata, and Anmol.Sarita.Bahl for kind words of encouragement and great suggestions.