Why Great Stories Fail on Medium and What You Can Do About it?Here are some surprising things I found while analyzing the Medium stories from 2018Ganes KesariBlockedUnblockFollowFollowingDec 28What makes an article go viral on Medium? Which ones do people actually read? What drives them to click the precious clap button, and when do they fall in love with it?These are questions on the minds of every writer. We’ll answer them here. We will make the most of the sparse but useful stats that Medium has to offer. We’ll then enrich it with insights from a million Medium stories published over a year, pulled together from public story cards by Harrison Jansma.This is also a reflection of my year on Medium, highlighting the stories that stood out and identifying some key takeaways on leveling up, as a writer.My 2018 on MediumWhile I’ve been blogging for some years, my journey with publications started earlier this year. I’ve published all of my 14 data science articles on TDS, almost on a monthly basis. A huge thanks to the TDS team for building this community and enabling access to such a quality reader base.My posts target 3 kinds of audience: a) how-to articles to help those aspiring a career in data science, b) explainers that demystify the techniques and methodologies for data science practitioners, c) articles that simplify data science for the non-tech and business audience.My 14 articles have garnered a total view count of 86,000+ (~120K including RSS views, republications and language translations) and 10,000+ claps..That’s an average of 6000+ views per post with a read ratio of ~40%.Summary of my stories on TDSIn comparison, Harrison found from his analysis that most Medium articles get fewer than 10 claps, and the top 1% receive more than 2000 claps..That’s about 2700 authors across all of Medium who’ve done that..Consistency is rarer since less than 0.3% of these highly clapped stories have repeat authors.While my top-viewed story ‘4 Ways to fail a data scientist job interview’ falls in this ballpark, it misses the cutoff for stories tagged data science..So, that’s something to target, along with the consistency.What makes people click on your story?A number of factors influence the exposure and view count of a story..While the publication’s reach, posting day, time and topic have a say, the clincher is the post title.“Unless your headline sells your product, you have wasted 90 percent of your money.“ — David OgilvyI experiment a fair bit with post titles, but I always dread that final phase of refining the headline before hitting ‘publish’..It’s a frantic analysis paralysis running late into the night, that eats up an hour, and leaves a bad aftertaste.Here’s what seems to work for my stories, but I’m still learning the ropes here.The media world has taught us that bad news sells, and titles benefit from a tinge of negativity (‘fail’, ‘frustrate’, ‘struggle’)..Listicles work and I use them periodically (‘4 ways’, ‘5 reasons’, ‘3 stages’)..Emotional and uncommon words bring in charm and appeal to a drab headline (‘secret’, ‘craze’, ‘real reason’).What’s the recipe for a great title?While my posts don’t have the volume, the million Medium stories offer some interesting insights in this area..Analyzing articles earning 2000+ claps with tags related to data science, here’s what I found from their titles.Not surprisingly, most titles had the keywords ‘data science’, ‘AI’, ‘machine learning’, ‘neural networks’, ‘data scientist’..Keeping these aside, we end up with this word cloud:Wordle.netMost titles are structured as questions ‘how’, ‘why’, ‘what’ and they engage well with users..These posts double up as explainers and use words such as ‘introduction’, ‘understanding’, ‘learn’, ‘guide’, ‘explained’, ‘tutorial’, ‘beginners’.The listicles stand out and it’s interesting to see the popular numbers: 1–6, 10, 15.. More details