This Headline Analyser Identifies the Most Profitable Articles

If you are going to use a headline analyser for your Medium articles— make sure it is on the money

You just wrote the most informative article in the world. Your prize-winning journalism could propel your writing career to fantastic heights. Fame and fortune await.

But, if your headline doesn’t attract any readers — what’s the point?

Some top bloggers argue you should spend as much as 80% of your time on headlines. If your audience doesn’t get drawn in by your headline, your content counts for nothing.

What Follows is Medium-Centric

I analysed the headline analysers. I looked at ten of my best-performing articles over the last year and put them through six headline analysers.

Each headline analyser provides an overall score. For example, my all-time best-performing article on Medium is:

I put the headline through six headline analysers. Results were mixed. Scores for this headline ranged from 39 to 73 out of 100.

I haven’t included this article in my analysis because it is somewhat of an outlier.

Published in December 2020, it had a slow start. Day one, it had 56 views, and despite it being a 9-minute read; it racked up only 20 minutes of reading time.

It bumbled along, not doing very much until the end of March when it took off, peaking at 558 views and 21 hours of reading time.

It tailed away and had another spurt all through April, peaking at 1.4k internal views and 42 hours of reading time, earning me $98 for one day.

It continued to do well in May and then hibernated until July 26, when suddenly it woke up with 10k external views.

To date, it has 961 hours of member reading time.

Nobody would have guessed this from the headline analyser feedback. I make this point early — headline analysers are not the be-all and end-all.

However, they can help you avoid the tumbleweed.

A compelling headline sells your idea. People won’t click on your articles unless you spark interest. So how do you know if it is any good? This is where headline analyser tools can help.

Which Headline Analyser is For You

Headline Analyser tools use their algorithms to assess your headline by comparing it to their database of already successful headlines.

Their formulas gauge niche metrics, readability ratings, and SEO ratings. Some also assess tone and sentiment.

Here's the rub — not all headline analysers are created equal.

Algorithms vary. Results can confuse. What might work on one platform could fail spectacularly on the next.

I picked six sites that offer free analysis. Some offer a paid version upgrade for added features.

1. Sharethrough Headline Analyzer

The Sharethrough Headline Analyzer has one of the nicer interfaces. It has a colour-coded overall score and breaks this down into scores for engagement and impression.

A handy feature is it keeps a list of your last 10 headlines so it is easy to compare previous scores.

2. CoSchedule Headline Analyzer

The CoSchedule Headline Analyzer also has a colour-coded overall score. It has a handy version history and offers tips on what types of words will improve your headline.

It also highlights word count, character count, headline type, reading grade, sentiment, clarity, and skimmability.

Whilst free to use, you must submit your name, email address, and some other personal information for access.

3. Capitalize My Title

Capitalize My Title has an overall headline score, readability score, SEO score, and sentiment score. There is advice on how to improve on these.

It has additional functions for capitalising headlines or changing from lowercase to uppercase and vice versa. This uses the same software as Medium’s article title capitalization tool.

4. Monster Insights

Has a list of power words, emotion words, and uncommon words to help you along your way.

5. Isitwp

Has a grid format with some limited analysis.

6. Aioseo

Has a score history and some useful tips for improving your headline.

№4, 5, and 6 all use the same algorithm. No matter what headline you put into these last three, they come up with the same score.

Their interfaces differ, so it will be a matter of preference.

Analysis of Results

These are my ten best-performing articles in the last year, in order of reading time.

1. Bitcoin Has One Big Flaw and You Can’t Deny It
2. I Was Wrong — This is Actually the Most Whackadoodle American Shit Ever
3. Fellow Time-Traveller Reveals the Big News to Come in 2022
4. This Remarkable Thing Your Brain Does When You Have Purpose in Your Life
5. It’s Whackadoodle Over There — This is the Most American Shit Ever
6. This Muscle in Your Calf Can Do Big Things to Improve the Metabolic Health of the Rest of Your Body
7. The Enduring Power of a Good F**k
8. I’m Never Buying Another Candle Again
9. This Is the Real Reason The Dollar In Your Pocket Will Be Worth 20 Cents in 3 Years
10. Half of All Traditionally Published Books Sell Fewer Than 12 Copies

It is of note that views, claps, and comments vary greatly.

№1, for example, received 796 claps from 69 people. I also received 41 comments — some not so complimentary because my article didn’t align with their views.

(There might be a lesson in that — controversy makes money.)

№10 received 3492 claps from 126 people, and 68 rather pleasant comments.

№1 earned four times as much as №10.

Earning money on Medium is all about reading time. The following analysis could also be done to compare claps or comments — if that metric is important to you.

I compared their ability to identify reading time because reading time = money earned.

I put the headlines through each headline analyser and noted the score. For example:

Article №1

Sharethrough score = 70

CoSchedule score = 65

Capitalize My Title score = 51

Monster Insights + Aiose + Isitwp score = 87

Article №10

Sharethrough score = 68

CoSchedule score = 67

Capitalize My Title score = 38

Monster Insights + Aiose + Isitwp score = 42

Since the sites use different algorithms, any comparisons between the numbers are effectively useless.

What I needed to do was compare how each site rated the top-paying articles and compare them to poorly-paying articles. I also chose ten titles that had all earned less than $5.

1. 3 Things People Share to Make Them Look Clever and Cool
2. Human Chain Helps Bookshop Move to New Premises — The Love of Books Knows No Bounds
3. Am I the Best Writer on This Platform ? — Um… No
4. That Sounds Like an Excuse to Me
5. Ever Wondered Why the Idiot Next to You Gets Promoted and You Don’t
6. If Anyone Was Going To End Up Addicted To Gambling
7. That Creepy Feeling Someone Is Looking At You
8. The Highly Questionable Tricks of the Uber-Rich to Avoid Paying Taxes
9. It Only Takes One Bad Decision to Blow Your Leadership Credibility
10. You Can Find Notorious Strumpets and Dangerous Girls in This Library

I totted up the total score and divided this by the total score for the top ten earners.

Sharethrough score = 95%

CoSchedule score = 105%

Capitalize My Title score = 107%

Monster Insights + Aiose + Isitwp score = 89%

What all this means

Overall, Capitalize My Title, proved worst for identifying the article headlines that had the most read time. Its algorithm gave a higher average rating to headlines that earned peanuts, compared to my top-earning articles.

CoSchedule didn’t fare much better. It, too, gave higher average ratings for the poor earners than the top earners.

Sharethrough was slightly better at identifying the higher-paying articles and Monster Insights + Aiose + Isitwp proved to be the best.

Another Important Comparison

A headline analyser is only as good as its algorithm. The headline:

“Edge Warn Soar a Really Fire Awesome Easy Love to Staggering”

Doesn’t make sense, yet, it scored an overall 88 on Sharethrough and a stonking 93 on Monster Insights. Capitalize my title wasn’t so easily fooled. It gave the headline an overall rating of 43.

So how do they ALL compare when tested with bad Medium headlines?

This is the most important metric.

I raked through Medium articles from 6 months to a year ago and found 10 articles with zero claps or comments. To save any embarrassment, I will keep these to myself.

I added up their scores on each of the headline analysers and divided them by my ten best-performing articles.

Sharethrough score = 65%

CoSchedule score = 56%

Capitalize My Title score = 108%

Monster Insights + Aiose + Isitwp score = 56%

This gives a more definite result. Overall, Captialize My Title still rated these higher than my best-performing articles.

Sharethrough had an improved differential. CoSchedule, Monster Insights, Aiose, and Isitwp scored best.

Caveat

This was a small sample based on Medium headlines. On other writing platforms, the results could be entirely different.

There is not a lot in it.

Headlines will also be subject to Medium’s model. It doesn’t matter if you write a crappy headline if your audience knows you well and knows what to expect from you. Loyal followers matter.

Medium’s relational model has to be taken into consideration. New writers on the platform won’t have a strong following, yet. They might nail the headline, but if nobody sees it, it won’t get read.

The point of the exercise is to highlight the importance of getting your title right. If you are struggling to attract people to your articles, there is something you can do about it.

Take your top ten earning articles and your ten worst-performing articles and put them through the same analysis with these headline tools. If there is a vast difference in reads — then that is something you can work on.

Your score isn’t so important once it gets above 60 on any of these headline analysers. Many other factors could influence whether or not your article goes viral.

However, if your headline scores below 50, rethink your copy.

PostScript

I put the heading for this article through the analysers and here are the results:

Sharethrough score = 69

CoSchedule score = 71

Capitalize My Title score = 42

Monster Insights + Aiose + Isitwp score = 93

Malky McEwan