POV: Hemingway says 83, your heart says zzz...

Three robot-perfect rewrites vs the originals that actually move humans—pick your side.

Day 134/100

 

Hey—It's Tim. 

Last night I ran my latest draft through the readability checker in Hemingway.
Score: “Grade 12 — Too hard.”
So I chopped, trimmed, smoothed. Grade 5.
Then I read it aloud.

It tasted like cold oatmeal.

Turns out--while I was busy pleasing an equation--I’d edited every ounce of personality out of the piece.

My voice was MIA, and the draft felt like it came from anyone (which means it came from no one).

“Is chasing 70-plus Flesch scores is flattening my voice and boring people?”

The problem with writing for robots

Don’t get me wrong, readability matters.

I spent five years as a structural-engineering drafter translating jargon for architects who just wanted the gist before coffee.

Short sentences kept us on schedule.

But readability is a floor, not a ceiling. If you’re contorting your copy so a fifth-grader can breeze through it, you’re probably stripping the nuance that makes adults lean in.

Flesch? Bless You.

There is method to the madness of reading score here. I feel that understanding how the formula actually works is good to know for when and where you need to break it.

The Flesch-Kincaid Reading Ease score is nothing mystical:

206.835 – 1.015 × (average words per sentence) – 84.6 × (average syllables per word)

• Fewer words per sentence → score climbs
• Fewer syllables per word → score climbs

Great for legal disclaimers and airline safety cards.
Terrible for lines that need flavour, rhythm, or edge.

Show your work

Here are three “Grade-5-approved” sentences… and the spicier originals they demolished.

Robot-friendly rewrite

The voice-driven original

“Our tool helps teams write faster.” 
(Flesch 83)

“Penfriend shaves hours off draft day, so Fridays feel like Fridays again.” (Flesch 54)

“The update may cause brief downtime.” 
(Flesch 79)

“Heads-up: servers nap at 2 a.m. Saturday while we bolt on the new turbo.” (Flesch 47)

“You should track your content ROI.” 
(Flesch 91)

“If your blog’s a money pit, let’s start digging up receipts.” (Flesch 58)


Lower scores, higher impact. Because the job of a sentence isn’t to ace a math quiz; it’s to make the reader feel something and do something.

So where’s the line?

  1. Clarity first; character close behind. 
    If a phrase muddies meaning, swap it. If it merely scares the formula, keep it.

  2. Read aloud. 
    Your ears will flag clunk before any algorithm does.

  3. Aim for “explain it to a smart friend,” 
    not “explain it to a smart fridge.”

My rule: I’ll bend a sentence for clarity, but never break my voice for a score.

Keep the colour in your copy, let the bots sweat the math, and remember:

Your best writing sounds like you spoke it on purpose.

✌️ Tim "Voice over scores" Hanson
CMO @Penfriend.ai

Same brain, different platforms: X, Threads, LinkedIn.

P.S. Why is it that the newsletters where I get to complain about traditional writing practices are also the newsletters that feel the most me?

Coincidence?

 

Penfriend.ai
Made by content marketers. Used by better ones.
 

What to do next

  • Share This Update: Know someone who’d benefit? Forward this newsletter to your content team.

  • Get your First 3 Articles FREE EVERY MONTH! We just dropped the biggest update we’ve ever done to Penfriend a few weeks ago. Tone matching with Echo, Hub and Spoke models with Clusters, and BoFu posts.

  • Let Us Do It For You: We have a DFY service where we build out your next 150 articles. Let us handle your 2025 content strategy for you.