I’ve watched teams make the same mistake over and over when they start building interactive tools with AI like calculators, quizzes, webpages, dashboards, and assessments.

They give up after their first attempt doesn’t match what they had in mind.

Here’s what works from my experience. Start with a simple prompt to get a baseline. See what AI creates, then iterate from there. Or go detailed upfront with specific user experience requirements.

Both approaches work. The mistake is expecting perfection on the first try and giving up when you don’t get it.

Imagine doing this: Work with AI to do deep research on a topic and build a working app in about 30 mins!

The calculator in this video (below) started with: “Create a simple business impact calculator for project management solutions based on the deep research report with metrics and benchmarks.” Then I continued to work with AI using plain English to make it look better and add features and messaging.

Play around with the calculator yourself (link in the comments). This one was created by Claude. But you can also create interactive apps with Gemini and ChatGPT.

AI builds in iterations. Sometimes a simple prompt gives you exactly what you need. Sometimes it gives you a starting point to refine. Either way, you’re building something interactive instead of just talking about it.

Tomorrow’s (June 26) newsletter shows you 10+ working examples and approaches to inspire what’s possible. I’ll show you HOW: the thinking, research, prompts, outputs from various AI models, and links to the shareable interactive apps so that you can try them.

Subscribe so you don’t miss it (link in the comments).

UPDATE (June 26)

In this newsletter (https://lnkd.in/gSQRrXfp), I share the complete execution playbook with 10+ examples of turning concepts into interactive experiences.

What interactive tools have you built or are you planning to build?

Get your ideas out of your head and turn them into something that others can easily understand and explore.

For the past 6 months, some of the most AI-forward GTM teams I work with have been building interactive tools instead of static presentations. Calculators instead of spreadsheets. Assessments instead of blog posts.

It was their secret weapon because only Claude could render working apps from simple plain language prompts.

That just changed.

ChatGPT and Gemini can now build and display interactive experiences. What was once a competitive edge for the few is now available to everyone.

Try this exact prompt with a specific article to see the transformation in action:

"Transform this article into an interactive 'Pick Your Challenge, Get Your Solution' tool. Ask users to select their main problem, then show them which solution from the article helps solve it, plus one action step to get started. Output the interactive app please."

Use any of these AI platforms (paid versions recommended and I got the best results using these models):

• ➡︎ Claude: Sonnet 4
• ➡︎ ChatGPT: o4-Mini-High
• ➡︎ Gemini: 2.5 Flash or 2.5 Pro

I tested this on a MarketingProfs article I wrote last year (see link in the comments) about AI use cases for CMOs. All three platforms built working interactive tools that let users select their specific challenge and get personalized recommendations.

All generally did a good job. But for me, Claude had the best design and pulled the most helpful insights. Gemini came in second and ChatGPT third. See my results in the comments.

Once you see how it works, adapt the approach to your own content. Each blog will need a different prompt based on its structure and topic.

We’re moving from telling prospects about solutions to letting them experience solutions. From explaining ROI to building calculators. From describing value to creating interactive proof.

This gives you a way to get ideas out of your head and into something others can explore. We all have brilliant concepts rattling around, but translating them into something people can understand and act on is hard.

AI democratizes software development. What used to require developers and weeks of back-and-forth now happens in one conversation. You can show your thinking instead of struggling to explain it.

UPDATE (June 26):

In this newsletter (https://lnkd.in/gSQRrXfp), I share the complete execution playbook with 10+ examples of turning concepts into interactive experiences.

Most people think the problem with AI is learning how to prompt. It’s not. It’s learning how to think.

See the comparison in the image. Both prompts use the same solid structure (in this case, the GRACE framework: Goal, Role, Action, Context, Example). Both have the same data.

But the thinking behind the prompt is fundamentally different.

➡︎ Approach 1 – “Help me reduce our churn rate.”

Result: Standard retention playbooks. Email sequences, loyalty programs, health scores. Useful, but expected.

➡︎ Approach 2 – “What if we’re thinking about churn all wrong? What if losing customers is revealing an opportunity?”

Result: A game-changing insight. The discovery that your “churned” customers are actually your most successful users outgrowing the product. Churn is reframed from a retention problem into a qualified lead signal for an enterprise upsell.

You get the most from AI when it can show what you’re missing about your own business.

This represents a fresh perspective that gives teams an edge.

Instead of focusing only on prompt engineering, the real advantage lies in better framing problems.

It doesn’t matter which prompt framework you use. What matters is the depth of the thinking before you even start writing.

I go deeper on how to develop this strategic thinking in my latest newsletter (link in comments: https://lnkd.in/d7jkBF7h).

What’s one core assumption your team isn’t questioning right now?