Master Storytellers

What separates master storytellers from the rest of us? I kind of stumbled on something neat and wanted to share it.

My friend Joe Lazer (Lazauskas) is a real storyteller, heck he even wrote a book on it, and a popular newsletter, and speaks on it.

I am a Pro Amateur, or a Pro-Am, self described term :D. Really like many of us, we all write for our job. I like to think I’m good.

But really… am I? I decided to find out.

For background I work at Nudge and we have been using ML to help predict attention & engagement on urls, including content.

We have an internal tool which deconstructs content, to analyze it and predict how well it would do. For example, we can predict say earned media. In this example, we looked at how well this press mention did.

This is saying that on average a reader would spend 45 seconds on that content, and for B2B that's 29% below the benchmark.

But as a by-product of this analysis, our system highlights, WHAT things are impacting engagement, through adding to the experience or taking away from it.

For example, it might say, you need to reduce buzzwords or industry lingo. Maybe your headline is too short etc. Or you have too many links.

So I thought, why don’t I compare. And I just took a recent newsletter Joe had written, and a random piece recently that we’d analyzed.

And this is what I wanted to share, because when I ran it, it wasn’t that surprising.

See below:

This is a pro writer. Nailed 4-5 things in the piece, which capture attention and hold them.

In the image above I have removed all the variable names. But each dot represents a specific variable that impacts attention.

The distance of the dots from the line, show each variables impact on attention. To the right, is adding to attention, to the left is taking away from it.

And the hierarchy from top to bottom shows the most important things impacting attention on that particular piece. In this case, Joe has nailed 4-5 things really well. Which deliver us with the end performance.

(As a competitive fellow, I'm sure he wants to remove those things taking away from attention, but they are minor, compared to the things which make it successful)

I was like, huh. That’s not that interesting.

It wasn’t until I looked at a regular piece, that you see the difference.

In this example, there are 6-7 things having a meaningful negative impact on attention, with only a couple adding attention. Uh oh.

My takeaway

Pro’s like Joe, side step the common mistakes the rest of us make. It’s like that design mantra, if you remove everything that is unnecessary, all you are left with is what is necessary. Making for a better product.

Or even an investing mantra, of let your winner runs, as they lift overall performance. Which is exactly what Joe has done. In my analogy, I've bought and sold too early.

That’s what Pro writers do, they just deliver what is necessary, whilst the rest of us, mix in a few mistakes. Super interesting! Basically they sidestep all the known potholes, and just hit the points which drive attention.

What does this mean for all the AI writers? Well if they are just writing, based off ‘input’, which they are. It means reverting to the norm - so output but maybe not the best output.

But with the right tools & datasets, they can produce better writing, or at least in theory. LLMs need their own feedback loop to evaluate if the output is good, and how to improve it. Neat huh.

This application has super interesting use cases:

• Compare your content vs someone one else’s.
• As a feedback loop to improve your own content.
• An input for AI to improve content.

We are already using these outputs for email, earned media, contextual advertising and naturally to improve content.

So if you’re interested in building on this too, happy to have a chat.

And if you have any content you’d like us to run, send it over, happy to do a couple of examples.

--

I also do a weekly newsletter at the intersection of advertising, media & content. You might enjoy it, can subscribe here. And see the most recent edition.