The two metrics everyone talks about are list building and conversions.
Opens don’t get discussed much any more, everyone knows that number’s unreliable.
But there’s one metric that matters more than either, and almost no one talks about it.
Drum roll.
Retention.
Here’s why.
Say your list churns 5% of subscribers a month. Cut that to 2.5%, and a year from now you’re carrying nearly twice as many active subscribers as if you’d left it alone.
Now boost your sign-ups by that same amount instead.
You get more names on the list, but if they churn at the same rate, you’re just refilling a leaky bucket faster. Retention compounds.
Sign-ups don’t.
And the subscribers who stick around are your best source of new ones.
They refer people.
They reply.
They buy.
Like all of us, you want subscribers who open your emails, reply to some of them, buy your products, and stay on your list for years.
You’ll still have to earn the sign-up.
Solve one problem for them, or run them through a quiz that lands on a useful conclusion.
Whatever it is, it needs to be worth more to them than their email address is to you.
Then your first email needs to be a quality filter, not the soppy “thanks for subscribing” note that’s about 90% of what lands in people’s inboxes.
One thing most people get wrong about email: they optimise for open rates. Open rates are a vanity metric, revenue per subscriber is the number that matters. Over the next four emails, I’m going to show you the three levers that move that number. Hit reply and tell me: what’s your current open rate? I read every response.
If a new subscriber doesn’t reply within 48 hours, they’re not a good fit for your list.
Flag them for re-engagement or removal in 30 days.
Most autoresponders charge per subscriber, so clearing the decks fast saves you money.
And in my experience, subscribers who don’t respond fast rarely respond at all.
A responsive list of 500 beats a non-responsive list of 5,000, every time.
There’s one way I know to build a list like that: fewer people, every one of them worth having.
The 1970s Stanford marshmallow experiment did happen, and early follow‑ups did find that kids who waited longer tended (on average) to have better later outcomes like test scores and BMI.
The 1970s Stanford marshmallow experiment did happen, and early follow‑ups did find that kids who waited longer tended (on average) to have better later outcomes like test scores and BMI.
But the test itself was flawed, and the results misleading.
Because the test was carried out with a very small, unusually privileged group of children at Stanford’s nursery school.
Later studies much larger and more diverse samples found that the “marshmallow effect” is much weaker and mostly explained by background factors such as family income, parental education, and home stability.
A 2018 replication (about 900 kids) showed that once you account for socioeconomic background and parenting, the marshmallow waiting time predicts very little about life outcomes by age 15.
A follow-up in 2020 of the original Bing Nursery School kids into their 40s found no meaningful relationship between how long they waited and adult outcomes like wealth, education, or health.
What this means in practical terms is that while delayed gratification can be important, it is not a reliable indicator of you or your child’s future.
Matt Furey found it, filmed himself demonstrating the moves, and turned it into a video course.
He’s still selling it.
Right now it’s $297, marked down from $497, and going by his own sales page, he’s kept some version of that price structure running for years, nudging it up as demand grows.
Project Gutenberg (gutenberg.org), and the Internet Archive (archive.org) have thousands of these, to download and use however you want.
Grab two or three on the same subject, feed them into notebooklm.google.com, and ask it to build a course, a book, a video series, whatever you’re picturing.
NotebookLM will do some of its own research and update the material as it goes.
That’s useful.
It’s also the part you can’t skip past.
Proofread everything it gives you.
Rewrite the awkward parts.
Check every fact it states like it might be wrong, because sometimes it is.
What you ship needs to be worth someone else’s money, that’s your job, not the AI’s.
It pays to remember that AI is neither artificial nor intelligent.
It’s just very good at making you think it is.
Enjoy the books either way.
Regards, Brent
P.S. If turning old books into new content sounds worth trying but “write good prompts” is the part you’d dread, that’s what the AI Prompt Guide is for: https://link.ckv.to/prompt-guide.
But we’ve all seen stats twisted so many ways to prove a point that we don’t trust those either.
Here’s a classic example of how that works (slightly simplified, but you’ll get the idea).
A drug company tests a new version against the old one. Old drug: 5% effective.
New drug: 10% effective.
So they market it as “100% more effective.”
Technically true. Also misleading.
The best way to get someone to buy, with no argument needed, is a free trial.
Either it works and they buy, or it doesn’t and they don’t. Simple.
That works fine for a lot of digital products.
Doesn’t work so well for physical ones you’ve already shipped out of town.
But there’s a way to do it with digital products too, PDFs and videos specifically, where access gets automatically revoked if they don’t buy after the trial, or they get refunded if they already paid.
It’s a locking system built for exactly that: you set the trial window, it handles the cutoff or the refund, no manual policing on your end.
It seems that every man and his dog is creating faster and faster.
Products, training, courses, etc. probably mostly AI created.
Not that being AI created is necessarily a bad thing, but…
It’s not necessarily a good thing either.
The quality of the AI creation will inevitably degrade.
Because almost all the AI tools train on everything that’s been dumped online, and so much of it is now AI slop the AI tools cannot help getting worse.
With this PDF you’ll build your own prompt assistant, one that already knows your niche and your voice, so you’re not starting from a blank page every time.
But, as our buddy Tommy said, goals without action are pointless.
Well, he didn’t say it quite like that, but that’s what he meant.
Your goals should excite you.
They should not intimidate you because most people over-estimate what they can get done in a week, and under-estimate what they can get done in a year.
So set that BHAG, then break it down into steps.
What has to happen before the goal is reached?
That’s a sub-goal.
What has to happen before that sub-goal is reached?
That’s another sub-goal.
Keep asking that question, keep breaking it down until you’ve got to the what has to happen today as the first action step.
Because in a lot of cases, the “person who delivers” can just be you, with the right prompts.
Writing, content outlines, social posts, scripts, even first drafts of designs and visuals.
AI can do a lot of that work now, if you know how to ask for it properly.
Which is the whole reason I put together my AI Prompt Guide.
It’s not a generic “how to use AI” PDF.
It’s built specifically for getting output you can hand to the client with no rewrites, fast, without wasting twenty attempts getting the wording right.
If you’re going to explore drop servicing, or you’re already doing some kind of freelance or service work, this closes the gap between “I found a client” and “now I have to actually deliver this.”
