This week I had a rather bewildering experience creating an absurd amount of marketing assets for the StoryLearning App with AI.
I conservatively estimate that the stuff I got done this week alone would have cost me $10-15k in freelancer fees, plus WEEKS of agonising back and forth.
Instead…
I just did it all myself, with the help of Claude Code.
So that’s what I’ll tell you about here.
| Sidenote – What you’re reading here is me “live blogging” the growth of my new StoryLearning App. If you’re interested in the chronology of it all, you can read back issues on the blog here. |
Now that my time has freed up a bit after the summer and finishing V1 of my new book, I decided to devote a chunk of time to “low hanging marketing fruit”.
What’s everything I own and have easy access to online that I can add to, change, or otherwise repurpose in order to get exposure for the app?
My starter list:
- Raw traffic
- Stuff that helps AEO (i.e. discovery in the LLMs)
- Appearing in bottom-of-funnel / purchase intent content
- Audiences for retargeting
Now, over the last couple of years, I’ve been building up quite an arsenal of AI marketing skills, but the truth is I haven’t really found opportunities to use a great deal of it.much.
For example, in my personal brand (here), I have a blanket no-AI policy, so nothing you ever read from me here will be written by AI, other than potentially a purchase confirmation email or something transactional like that.
And then inside StoryLearning, as an established business, most of the AI things we could do fall more into the “optimisation” camp than really exciting new opportunities.
(Which I think is the case for the vast majority of businesses, hence why so many people are just wasting their time with AI.)
But…
With a brand new product to play with, I suddenly found ways to deploy these skills to make a LOT of progress very, very fast:
- Blog post writing
- SEO research
- Meta ad creative
- Email writing
- Deep research
So, last week, like a rabbit in spring, I sat down and got busy deploying all of these skills in quick succession, creating a dazzling array of Claude agents running simultaneously and resulted in most of my week spent staring at my laptop hitting the “Accept” button.
Fun?
Not so much.
Productive?
Just a bit!
First, I researched, wrote and published 10 new blog articles for the website focusing on competitor product searches (bottom of funnel, high purchase intent).
Next, revised 20 existing posts on the website to include contextual mentions of the StoryLearning App.
These two are designed to divert existing traffic to the app, and to create targeted material for ranking in LLMs.
Then, I started thinking about all of the existing traffic that we are getting to the app (from sources that I have yet to properly understand – more on that in the coming weeks) and what we could do with it.
It breaks down into the following buckets:
- 100,000s of people who land on the general storylearning.com website each week (but don’t go to the app)
- People who land on the app sales page specifically (but don’t buy)
- People who create an account (but don’t buy.)
All of these people can be considered warm prospects and can be retargeted in different ways.
For example, those who land on the general storylearning website can be sent an ad that just generally raises awareness of the app.
“Hey, did you know we had this thing? “
People who land on the storylearning app sales page but don’t create an account can be sent an ad saying,
“Try it today.”
And then people who create an account but don’t buy can be shown an ad saying,
“Hey, finish your purchase. “
So I put my Claude skills to work and set up all of the following.
- Setting up end-to-end GA4 tracking, with Tag Manager fully set up to track different events, such as website visits, account creation and purchase
- FB events to track the same
- Custom audiences for people who have landed on the website, landed on the sales page, created an account, but not bought, etc
- And lastly, a full suite of differentiated ads to run to these audiences
As if that wasn’t enough for one week, I then realised that we didn’t have a good abandon cart sequence set up.
So I got Claude to research best practice for abandoned cart sequences on consumer education apps, propose a 4-day email sequence based on other emails and content I’ve already written, and then queue up all the emails to be set up in a kit automation.
Result:
Everyone who creates an account, but doesn’t buy, will be sent a four-day countdown sequence with a 10% discount code that expires after 48 hours.
(Shh… keep this a secret please)
.
.
.
Did I mention I did all this myself… part time… in 4 days?
In fact, I wrapped all of this up by Thursday afternoon and still had Friday left to go…
but I was so dazzled by the amount I’d managed to get done this week that I ended up removing my ass from the chair, closing my laptop, and spending most of today enjoying the culinary delights of Hobbitshire.
Not every day has to be a school day, after all.
Now, in amongst all this craziness, there is also one other major project that I started and made headway on this week…
But it involves a different type of audience…
And I’m not ready to talk about that yet…
So consider yourself cliffhanger-ed!
Anyway…
I hope this gives you an insight into the sort of stuff that I’m working on, and if any of these AI skills or marketing processes are of any interest, feel free to reply to me and I’ll try to drill down onto them in future emails.
In other news…
This week has been another great week for sales in so much as we are making more sales than we are losing monthly customers to churn.
Which is great, especially considering the fact that we’ve discovered we have a monstrously high rate of churn on the monthly plans.
I won’t go into that here, but it’s safe to say a lot of my attention is going to be on this in the rest of the year.
Anyway, I’m now off to catch a plane to Tenerife for our biannual Creator Empires mastermind.
We’ve got 20-plus creators descending into a villa in Tenerife for a week of food, sightseeing, and cocktails.
We might talk about business at some point too, but let’s not get ahead of ourselves.
There’s more to life than Claude Code.
Namaste,
Olly
