hey Joshua - I've dealt with this exact CF7 webhook issue before. the plugin saves the URL to wp_options and some caching plugins or WP updates wipe it back to default. quickest fix: ditch the CF7 webhook plugin entirely and use WP Webhooks or just add a small functions.php snippet that fires on wpcf7_mail_sent hook to hit your n8n production URL directly via wp_remote_post(). way more reliable and you keep the email notifications working too since you're not overriding CF7's mail action. if you want I can hop on a quick call and walk you through it - I've built this exact WordPress → n8n → CRM pipeline for a few clients. DM me if you need a hand

@Enzo Kamga anytime brother, wish you the best!

@Enzo Kamga sure, feel free to!

This comes up all the time. Here's my approach: First, I try to keep the automation layer separate from the client's existing stack whenever possible. n8n sits in the middle and connects to pretty much anything via APIs or webhooks. So even if they're using some CRM I've never touched, I just need to figure out the API docs - not become an expert in the whole platform. For unfamiliar tools, yeah I'll create a free account to test the API endpoints before touching the client's setup. Usually takes 30-60 minutes to understand how their system handles data. The n8n HTTP Request node is your best friend here because it works with literally any API. A few things that have saved me: - Always ask the client for API documentation or admin access upfront. Most platforms have decent docs. - - Use Postman or the n8n HTTP node to test API calls before building the full workflow - - Record a Loom of yourself walking through their current process so you understand the flow before automating it - - If their stack is truly obscure, be upfront about the learning curve and factor it into your pricing The biggest mistake is pretending you know a tool you don't. Clients respect honesty way more than fake confidence. I'll straight up say "I haven't worked with [tool X] before but the integration approach is the same - give me a day to map out your API and I'll have a plan." What specific stacks are you running into?

Honest take: the tool-hopping is the trap, not the solution. I went through the same phase. Here's what actually moved the needle for me: I stopped learning tools and started solving problems. Sounds obvious but there's a real difference. When you're "learning n8n" you build tutorial projects that go nowhere. When you're solving a specific business problem (like "this dental clinic loses 30% of calls after hours"), you learn exactly the parts of n8n you need AND you have something to sell. To answer your three questions directly: No, you're not wasting time on n8n. It's the backbone of most AI automation agencies right now. New tools come and go but n8n + a good AI model handles 90% of what clients actually need. Stick with one system. Go deep on n8n. You can always bolt on other tools later (voice agents, specific AI models) but n8n is your orchestration layer for everything. If I had limited time starting today: pick ONE niche (restaurants, dentists, real estate, whatever). Build ONE automation that solves their biggest pain point. Then reach out to 20 businesses in that niche and offer to set it up for free in exchange for a testimonial. You'll learn more from one real deployment than 6 months of tutorials. The people making money in this space aren't the ones who know every tool. They're the ones who can walk into a business and say "I'll fix this specific problem for you." What niche interests you most?

Another approach worth considering: instead of fighting with Figma's API directly, use n8n to generate the landing page as actual HTML/CSS. Here's why: You can have an AI node (Claude or GPT) generate a full landing page in HTML based on a brief, then n8n can automatically deploy it to something like Vercel or Netlify via their APIs. The output is a live, editable page - not just a mockup. If you specifically need Figma-editable output for client review, the Figma REST API + n8n HTTP request nodes work but it's tedious. You'd need to create a template file in Figma with named components, then use the API to duplicate and swap text/images. It works but takes a while to set up properly. The fastest pipeline I've seen for Meta ads specifically: n8n triggers → AI generates copy variations + image prompts → image generation API (DALL-E or Flux) → outputs directly to an ad-ready format. Skip the design tool entirely for performance ads. What's the end goal - are these for client presentations/approval, or are they going straight to live deployment?