Activity
Mon
Wed
Fri
Sun
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
What is this?
Less
More
7 contributions to SaaS Ads
0 likes • 14d
For the first test, keep the offer, audience and landing page fixed so the edit is the main variable. Watch the opening hold and click-through, but use add-to-cart and purchase cost to decide whether it brought qualified traffic. If you change the hook and the edit style together, it will be hard to tell what helped. I'd compare a control with one version that changes just one element, then check whether people drop between landing-page views and add-to-cart.
Welcome! Introduce yourself + share a pic of your workspace 🎉
Let's get to know each other! Comment below sharing where you are in the world, a photo of your workspace, and something you like to do for fun. 😊
1 like • 18d
Hi, I’m Noah, based in Germany. I work on Shopify CRO and on-site AI chat, using shopper questions and funnel data to improve product pages. I try to keep ad and landing-page tests clean by holding the offer and audience steady, then judging the change on purchases or qualified actions rather than engagement alone. Glad to join and learn how others are testing creative and landing-page fit.
I Replaced a $1,000 Copywriting Agency With One YouTube Video
I paid an agency $1,000 to audit my landing page copy once — and regretted it. Here's what I do instead now, for free. The trick: find the longest, most credible YouTube video you can on landing page copywriting, turn it into a Claude skill, then paste your own landing page URL in and get an instant, detailed audit. In this clip: - Why I stopped paying for copywriting audits - The exact process for turning any YouTube video into a reusable AI skill - A real audit example, run on my own landing page
I Replaced a $1,000 Copywriting Agency With One YouTube Video
0 likes • 18d
An AI audit can surface useful hypotheses, but I'd treat each copy recommendation as a test rather than adopting a full rewrite. Pull the main promise, proof and friction points from customer questions and support tickets, then change one hero element at a time. Track hero-to-primary-action clicks alongside qualified sign-ups or purchases, since more clicks can still mean lower-intent visitors. Save the control and variant, and compare them over the same traffic sources before deciding the change worked.
0 likes • 18d
Mapping each ad format to one primary job makes testing much cleaner. I'd keep the offer, audience and landing page fixed, then vary one creative format or framing at a time, with ad names that record the intended job. Use hold rate and CTR to diagnose attention, but pick the winner on cost per qualified lead or purchase and downstream revenue. If the platform sends very uneven spend, use a randomized A/B test rather than treating delivery as proof that one creative won.
Aug 14 - update
I don't post these often, but I want to start sharing my Meta ads for upnova.io in real time, what's working, what's not, so we can learn from each other. My old strategy stopped working after the algorithm shift, and I'm recalibrating. What I cover - Current numbers: ~$400/lead in the US (targeting $300). Also running Australia since CPM is half, letting me test faster and riskier there. - Transparency: All my ad creative is public, search "Upnova" in Facebook's Ad Library. - How I judge ads: CTR, cost-per-click, CPM, hook/hold rate as a pulse check, but cost-per-result is what actually matters. - The problem: Most of my content doesn't mention the offer at all. "Switchcraft" got great engagement, zero leads, killing it. Only 3 of ~15 ads actually address the offer/pain/solution, and those are the ones converting. - The fix: New structure for every ad: Awareness → Reinforcement → Echo → Decision → Proof → Objection handling. - Format: Mixing video (best performer) and images (cheap for hook-testing). - My tools: Custom dashboard (BB plugin, now an Electron app) tracking spend/CTR/CPC/CPM/conversions per ad, plus a funnel dashboard, big bounce drop after the hero section, ~20% form-open-to-submit rate. - Bonus tip: Turning ad-strategy YouTubers (e.g., Ben Heath) into Claude skills so I can query their advice anytime.
Aug 14 - update
1 like • 18d
Your notes point to two different leaks: message match at the hero, then friction after the form opens. First, use the ads that clearly name the offer or pain and A/B test a landing-page headline that repeats that promise with one piece of proof, tracking hero-to-form-start rate. For the roughly 20% form-open-to-submit rate, segment by device and test removing one nonessential field; track booked calls per landing-page view and cost per booked call, not only leads. Check lead quality by market before treating Australia's cheaper CPM as a win.
1-7 of 7
Noah Müller
1
2 points to level up
@noah-muller-2497
Blogger of Golf and Self Improvment

Active 1d ago
Joined Oct 5, 2026