How To Write A Good Blog Post Title

Stop following bad “How To” guides

Who cares about titles?

You, probably. Most writers think they should. And I get why:

“According to Copyblogger, 80% of your visitors will read your headline — but only 20% will go on to finish the article.”

So, titles matter. But: we care about — and are looking at — irrelevant things.

WE’RE LOOKING AT THE NUMBERS ALL WRONG

I’m amazed at how many people slap together averages of top posts and call it “analysis,” actually thinking this is sufficient, let alone “significant” on any intelligent level.

First: top post “averages” say something about them, but nothing about how they’re different than the low performers.

Without comparing, you can’t know anything significant. You have to look at the entire data set for any kind of meaningful insight.

Second: averages can mean absolutely nothing. Just because you can calculate an average (and god, do we freaking love to…) doesn’t mean it matters.

I could tell you the average name length of best-selling authors or athletes, but while that may be interesting, it‘s not what actually makes them good.

Quit with the “averages.” They’re fun, but illogical.

STATISTICAL ANALYSIS

I ran statistical correlation tests on the titles of my 470+ posts to date. (Feel free to skip this part if you don’t care about the details.)

Model used: Pearson correlation coefficient

It measures the linear correlation and covariance between two variables, and yields a value between +1 and −1.

  • A score of ±1 means total correlation (direct or inverse)
  • A score of 0 means there is no correlation

The closer to ±1, the more correlated. The closer to 0, the less.

An Example:
I ran the coefficient between “Reads” and “Views” (obviously very correlated)… and, yeah, it came out to 0.977 (p-value 0.0.), i.e., “fully correlated.”

Which, duh, makes sense. “Reads:Views” is our (actually correlated) control.

p-value
Probability value represents “the probability of the significance” of results.

Statistically: a small p-value (often ≤0.05) indicates strong evidence against the null hypothesis, so it’s rejected. A large p-value (>0.05) indicates weak evidence against the null hypothesis, so it’s not rejected.

Layman’s terms: <0.05 = the PCC “matters” / >0.05 = it may not

YMMV on PCC and p-values, but probably not much…

Read More Here

Leave A Comment

Your email address will not be published. Required fields are marked *

Open chat
Need help?