September 2025 wrap-up
What have I been doing and learning this month? This blog post triples as my notes, a status update, and a way to share the things I found interesting this month.
Highlights
I guess the main highlight is that I actually resumed attending webinars and doing these write-ups! The n8n contract is ongoing, and my plans for the rest of the year are starting to crystallize (a little work, a little learning).
Watching
The Marketing Meetup: How to approach GEO in 2025 - Laura Ingram
- GEO: generative engine optimisation.
- Nobody is a GEO expert yet.
- 48% of shoppers use AI in their search.
- Might be sponsored responses coming.
- Now is a good time to experiment.
- AI increases zero-click. Zero-click searches are when users find the information they need directly on the search engine results page (SERP) without clicking through to a website.
- Potentially fewer clicks, but users who do click through from AI search are "warm".
Main advice:
You should be:
- Specific: ideally AI should be able to quote you directly.
- Unique(ly valuable): proprietory reports; interactive calculators; anything your competitors don't do.
- Credible: very important! Appear in news sources and industry press; organic Reddit content; YouTube, socials etc.
- Human: use normal natural language
- Brand-led: "
is a . . . " not "We are . . . " - Informative: as always, genuinely helpful content does well.
- New: AI prioritises newer content.
- Structured: headings, lists, scannable information.
From the Q&A:
- How to test: be aware that AI will be personalising your results. There are tools you can use to check which AI tools are mentioning you (including some of the existing SEO tools like Ahrefs).
Comments
- Polished and clear talk.
- Some useful tips. Immediately went and did a pass for brand-led language on my own site.
- Not super-sold on the idea we don't even need a concept-level understanding of generative AI.
- Staggered she seems to be simply trusting the info returned by AI.
- Also quite surprised she was willing to make sweeping industry predictions.
The Marketing Meetup: How to use AI to help create great content - Ross Simmonds
Ross is excited about AI because it equalises the ability to tap into your own creativity and reach more people. It lowers barriers to access to build/create your ideas. He recognises the concern for arts and crafts, but points out we can't go back. He thinks there's never been a better time to be a maker/builder/creator.
Thinks marketing in recent years has bought in too hard to creating content: too much content production, not enough customer obsession or understanding of buyer psychology and buying decisions. Too much content output, not enough research. We need to stop treating marketing like a production line. It should be a blend of the art and science of communication.
People are craving human connection more, especially with the rise of AI.
When we should use AI:
- Analysis
- Crafting comms
- Automate repetitive tasks
- Transcripts
- Rewriting contents
- Gut checks
How to keep your humanity in AI, and what not to use it for:
- Source material can be supported by AI, but shouldn't be replaced by AI.
- Distribution.ai: tool he's built that generates content types from a source. But he emphasises that the initial source shouldn't be AI. For example: the webinar should be a conversation between humans, but can then be used as source for LinkedIn post, email, blog post etc.
- The biggest risk of AI: we lose the skills to know what good really looks like, and how to create it. Side note: Thinking first, AI second is relevant here. You still have to learn skills, you still have to be ok being bad at things for a while so that you learn how to be good.
Recommended tools and cost-friendly tools:
- An LLM (ChatGPT, Claude, Gemini - pick one, lean into it, learn to use it well)
- Big fan of ElevenLabs
- Distribution.ai His own product ;-)
- HeyGen: generate an AI version of yourself. A bit scary (real-looking videos that aren't real . . . ). He anticipates legislation around AI video fraud.
- Suno: AI music generation.
-
Replit for creating mockups and building in-house tools.
Side note: his example of creating an in-house expense tool is funny, because now they have to host, secure, and maintain that . . . I'd probably recommend paying the $13k a year for the third-party product.
What's the number one mistake he sees marketers making with AI? Copy and paste. You can't just generate content and be done. You need to review and edit. Have checklists to review and elevate the content.
On environmental impact: we can't slow down progress. It's happening. But we can try to be our best selves and achieve while doing good.
Comments
- I'm not convinced that AI is great if you actually enjoy being a builder. But it does reduce barriers to generating output and implementing ideas.
- The chat was interesting. Quite a few people had concerns around both reliability and environmental impact.
Your Product Needs Words