Choosing an AI Business Idea to Fully Automate
This post collects together my research on AI business ideas with the aim of choosing one or two business ideas to pursue as part of my 100% AI Business 2025 challenge.
Whatever I build needs to be as easy as possible to automate with the latest AI tools; let’s jump in.
New here? Welcome! This is the journey of building a 100% automated AI business in 2025. You’re jumping in after we’ve already kicked things off, so you might want to catch up first.
Check out these key posts to get the full story—and don’t forget to subscribe for updates and exclusive perks:
Finding and Rating AI Business Ideas
I’ve run an eclectic range of businesses in my lifetime, but mostly they’ve been software businesses.
As an engineer at heart it always made sense for me to build tools with code.
But as we see AI stepping up faster and faster, my plan is to lay down my tools and start fully automated AI businesses, (that’s why I’m doing this challenge).
So how do you go about finding business models which are easiest to automate with AI?
Use AI.
… In the end I also wove in some of my own business ideas, but fundamentally this is an example of the nested power of AI.
Enter: AI Business Ideas directory
This is one of the pages in my new AI Business Ideas directory. I built each page using a combination of my own take and AI inference. Each page shows up in the ideas directory, and has an analysis of the complexity of automating that business model.
Aside: pSEO and Business Agents
You may have noticed that here I’m doing something we call pSEO (programatic SEO) – making a number of ‘semi-automated’ content pages.
For added fun here, I’ve been rewriting my code for this (with AI 😂) so that later it’ll be a natural function of an AI agent.
This challenge will see me build a lot of different tools, and AI pSEO will be key, so it feels great to get ahead of the curve a bit.
… but I digress
Choosing an AI Business Idea
So now I’ve had AI crank through my various ideas and categorise them by difficulty. This gives me a place to start.
I’m going to use my favourite idea shortlisting technique: Matrixing.
Here’s what I’ve done:
- I put all the ideas in the ‘B/C’ grade complexity from my AI business ideas page into a spreadsheet matrix.
- I added columns reflecting my overall goals of the project.
- I added a column for “Woody Excitement”.
- I filled out values using my best judgement.
- I added a little weighting, then sorted by total value.
If you’re interested in how I did this, here’s the spreadsheet I made.
Here are the top 10 ideas:
| Idea | Complexity | Scale £10kpm | Value-add | Do no harm | Long Term | Interest | Total |
|---|---|---|---|---|---|---|---|
| Automated CRM System | 4 | 5 | 4 | 5 | 5 | 4 | 29.7 |
| Online Course Creator | 4 | 5 | 5 | 5 | 4 | 4 | 28.5 |
| Personalized Email Newsletter | 4 | 5 | 5 | 4 | 3 | 4 | 27.7 |
| AI-Powered Market Research | 4 | 5 | 3 | 5 | 4 | 4 | 27.7 |
| AI-Powered Job Board | 4 | 5 | 4 | 5 | 3 | 4 | 27.7 |
| Automated Meme Artist | 3 | 4 | 4 | 5 | 4 | 5 | 27.7 |
| AI-Powered Podcast Editor | 4 | 5 | 4 | 4 | 5 | 3 | 27.4 |
| Infinite Text Adventure (game) | 3 | 4 | 5 | 5 | 4 | 4 | 27.4 |
| pSEO as a Service | 4 | 5 | 5 | 3 | 3 | 4 | 26.7 |
| Directory Website(s) | 4 | 5 | 4 | 4 | 3 | 4 | 26.7 |
… now as always with these ‘idea matrices’ the outcomes can surprise you.
I look at this list with dubious excitement. Some of these could work, I think, but I’m only certain on a few of them.
We all have to start somewhere, though.
Shortlisting AI Business Ideas
The above process gave me 10 ideas which could all probably work to some extent. To begin we’ll narrow these down to 3 and do some experimentation to see which is the lowest hanging fruit, if any.
My gut says, let’s start with:
- AI-Powered Market Research – It’s curious that the AI suggested this, because one of my small bets last year was NicheReport.io which is basically this (50% automated currently). It’s a no brainer to explore ‘finishing’ that business with AI.
- Personalised Email Newsletter – This one appeals to me because email seems to be the protocol that just won’t die. I also think there is lots of genuine value-add that can be generated here, for both the newsletter maker and the subscriber.
- Directories/Job-boards – I group these two into one because they amount to pretty much the same thing. I think of all of these, this one is probably the easiest automated, though with AI causing such radical shifts, I wonder how long we’ll have Google + Directory sites generally.
(I’ve also chosen these three because there is an amount of crossover in the AI agents required to achieve each.)
This challenge will only be achievable if I have some success with agents I can reuse.
I reserve the right to completely backtrack on this idea selection at any point 🙈.
Experiment Round 1: First Shortlisted AI Businesses
It’s finally time to start experimenting! Here’s what to expect from me in the coming weeks:
- Breaking down these business models into agents/workflows
- Testing existing AI tools which may work for these ideas
- Writing prompts & agent frameworks to wire all this together
What’s your take? Do you have any other good AI business ideas? Please do share in the comments if so, help me out!
P.S. I’ll be adding other AI business models to the list as I go, especially as I run tests on each and see how that pans out 😂.
Leave a Reply