From my desk: guides from real projects
These articles come from real projects — what worked, what I learned, and how you can apply it to your own content challenges.
Worried about using AI to write?
Here are five methods to keep your authentic voice while letting AI handle the heavy lifting — from feeding it facts to the “contrarian prompt” that catches weak spots before your readers do.
Key Highlights
- Five practical methods that work independently or together
- The “contrarian prompt” technique for catching weak spots
- Real examples from writing artist profiles for a local arts center
- Framework you can reuse for any content project
More to explore
Coffee and Claude time: Three prompt examples (and more)
A quick walkthrough of prompt approaches and when to use each one. Read on LinkedIn →
Highlights
- Three prompt examples in decreasing order of how much you bring to the table
- Why “coffee chat mode” beats command-line prompting
- The one addition that shifts AI from answering to collaborating
- Treat your AI assistant as a colleague, not a vending machine
How I built a lead capture system in 90 minutes
You don’t need to be a programmer or wait for perfect AI tools. This walkthrough shows how I built a working system to process business cards, create contacts, and draft follow-up emails — using Claude, Google Apps Script, and zero additional subscriptions. [Read on LinkedIn →]
Key Highlights
- Real example: from problem to working system in one session Four-part framework: AI extraction, human context, automation, hybrid handoffs
- Step-by-step guidance for building your own AI-augmented workflow
- No coding required — just collaboration with your AI assistant
Friday fun: Songwriting with Claude AI and Suno AI
What happens when you feed 160 pages of chicken observations to Claude and ask for Mother’s Day lyrics? A country rock anthem that nailed it on the first try — no 21-version iteration cycle required. Read on LinkedIn →
Highlights
- How rich context and clear constraints reduce AI iteration cycles
- Pairing complementary AI tools: Claude for narrative, Suno for music
- Why “one and done” worked here but not on other projects
- The human-at-the-beginning-of-the-loop principle
Tools worth trying
Free tools for checking how search engines and AI see your content
- HubSpot AI Search Grader: This free tool scores your website based on how well it’s optimized for AI-powered search. It gives clear suggestions across three pillars: content clarity, structure, and trustworthiness. Visit Tool →
- Google Rich Results Test: A must-have for checking if your content is eligible for enhanced AI-driven displays like FAQs, star ratings, or how-tos. Schema markup matters more than ever. Try It →
Content clarity and prompt tools
- Hemingway App: Quickly check your writing for clarity, passive voice, and complexity. Clean content is easier for both AI and humans to understand. This is ideal for rewrites before publishing. Use Hemingway →
- Siftly.ai: Get visibility into how LLMs interpret your content. Especially useful if you’re already using structured content and want to validate how it’s being summarized, scored, or indexed. (Free plan available) Explore Siftly →
Case Study: Radionix Website
From nostalgic storytelling to search-ready structure
This mockup shows two versions of the Radionix website for demonstration purposes.
Radionix is a fictional, family-run company making transistor radios since the 1960s. The “before” version captures the warmth and history of the brand, but — like many small business sites — it’s not optimized for modern searchability.
Old version: Rich story, low visibility
The legacy site is heavy on narrative and light on structure. While visitors get a vivid sense of Radionix’s origins, products, and values, much of the content is locked inside long paragraphs with few technical details, no product schema, and no Q&A-style answers.
Search engines and AI assistants can’t easily pull facts from this format, meaning Radionix could be invisible in AI-driven search results.
Revised version: Still authentic, now optimized
The updated site keeps the nostalgic storytelling but organizes product information into clear, structured sections. It adds:
- Consistent product spec tables
- FAQ modules written in question-and-answer format
- Descriptive image alt text
- Schema markup for Product, FAQ, and Organization
- Internal links that connect historical content with current models
The result? The Radionix story is still front and center. But now the site speaks fluently to both human visitors and AI-powered search engines, improving discoverability without losing personality.