How I work with AI
Human perspective, AI support. I’m Alastair, or Al to my friends. I’ve been using AI across various businesses for the past three years, and while I’ve found it incredibly useful, I’ve experienced plenty of its frustrations too.
My lightbulb moment on the limitations of AI automation: Yoast schema files
I’d lived through the common problems of AI trying to turn my content into generic slop, of it hallucinating then confidently presenting “facts”, but it wasn’t until I began to question why Google and AI platforms weren’t understanding one of my businesses that it really hit home.
Like many people, I knew I needed schema files, but didn’t know much about them, so I was happy to trust an automated facility on my hosting platform.
When I poked my head under the bonnet, though, I discovered that what the platform had presented was incorrect in places and missing lots of key information. So I researched what was actually required, then worked with AI to build out the missing information and code properly.
Within weeks, the business was being cited by AI platforms and I was starting to get leads through them. More importantly, my relationship with AI had changed from enthusiastic user to master inquisitor.
What AI support brings, and what human perspective transforms
AI, if properly directed, can do a lot of the heavy lifting, such as research and technical groundwork. Given my past three years of experience, I am naturally cautious and often have multiple platforms checking each other’s homework before my final pass.
How AI can support
- Analysing competitors and markets
- Investigating how Google and AI platforms interpret a sector
- Building schema and structured data following my input
- Comparing how different search environments handle the same business
- Identifying technical opportunities
- Speeding up research and fact-checking
- Testing how different AI systems currently understand a website
My human perspective role
- Understanding the founder and the business behind the website
- Prioritising the commercial content according to customer search data
- Identifying which customer reviews and testimonials to promote
- Finding the people stories that give a business a personality and build trust
- Creating content and layouts that work for the people using the website while giving search engines and AI platforms the information they need
- Final judgement on what gets published
Copywriting: the power and subtlety of words
AI is trained to write in a particular style, and it’s pretty obvious when we see it. Suddenly, there’s a credibility gap: how do we trust the person if we can’t trust the words?
A lot of the success I had over 30 years sourcing and marketing wine came from understanding that it wasn’t just about finding the right liquid. The story behind the wine, the winemaker and the region could be what influenced a buyer to list my wine rather than someone else’s.
That same emphasis on research, checking and getting the detail right has carried into my own writing. Four articles from my Alcohol Reduction Coach business have now been independently indexed by Google Scholar, and Scholar itself has started sending visitors to the site. It was never something I set out to achieve, but it is a useful indication of what that level of rigour can produce.
I bring the same added value thinking to website design. Two comparable businesses can offer much the same service, but the one that communicates its experience, personality and point of difference properly has a much better chance of being trusted and chosen.
Not sure what a technical term means? Check out our plain English glossary.
That is how I work with AI.
I use its speed and technical capability extensively, but I question it, check it and make the final decisions. Human perspective, AI support.