I've been doing discovery calls with businesses lately, mapping where AI can move the needle for them. My frame is dead simple. AI needs to do one of three things: create time, create value (save or make money), or provide knowledge.
We always find opportunities. Always.
But here's what I'm seeing more often: businesses that aren't ready for AI at all.
Not because they lack vision or budget. Because their operations are held together with digital duct tape and wishful thinking.
You can't automate chaos. Throwing an agent into a mess just creates more mess.
Before we can even think about AI, most businesses need to solve two unglamorous things:
Data organization and structure. Your information needs to live somewhere predictable.
Defined, repeatable SOPs. The core work that drives your business needs documented processes.
Things like:
Customer support running out of individual email accounts
Critical context shared only in meetings, never written down
Using memory instead of a CRM
Contracts and proposals scattered across random folders in SharePoint (oof) or Google Drive
Teams communicating across email, text, WhatsApp, and Slack or Teams (oof again) simultaneously
Sound familiar?
New segment: Reader Question
I get questions from readers occasionally. I’ll start answering some each week.
“I run a marketing agency with seven people. Some of us use ChatGPT to write emails, tidy up proposals etc. but that’s it.
I've tried running "AI idea workshops" to find other ways we can use it in our business, but it's difficult to get people engaged. Tips? “
I see and hear about this often. People orient to AI in two different ways. Here’s an example:
Some engineers use vibe-coding tools to experiment and see how crazy it is that you can prompt a proof-of-conept into existence. Other engineers will do the same only to search for errors to point out the shortcomings.
The prospects of using AI triggers fear and defensibility for a lot of people. A designer doesn’t want to experiment with the latest image models because they’re not sure if it means they’re putting themselves out of business.
You need these three things to drive engagement:
- examples to inspire
- shared learning
- setting expectations (clarify what the intention is, address concerns and fears)
Have a question? Send it here.
For smaller businesses, my go-to recommendation is building everything around Notion as the hub:
Set up basic company structure: Clients, Projects, SOPs, Team docs all in one place
Create data relations: Connect your client info to project timelines to invoices
Document the SOPs: How to use each part, who's responsible for what
Then you add digital workers to the mix.
What's great about Notion is it's flexible enough to grow with you but structured enough to keep AI from going sideways. You can start simple and add automation layers as you get comfortable.
PS: I'm not getting paid by Notion to say this. I'm just a fan of using it as the foundation for AI-enabled businesses.
I'm excited for the day when we can dump everything into our business's AI brain and it sorts itself out automatically.
We're not there yet.
So we do the unglamorous work first: organize the data, write down the processes, create the structure. Then AI becomes a force multiplier instead of a chaos generator.
If you're thinking about AI for your business, start with an audit of your current operations. Are your ducks in a row? If not, that's step zero.
(And if you’re not sure how to go about this yourself, you can book a free discovery call with my AI consultancy, Glassboks).
If you have ideas you want me to write about, hit reply!
Until next week,
Martin