←
Johann Van Tonder3,775 posts
Johann Van Tonder avatar

Johann Van Tonder

@chat_JVT

CRO since 2009 for Avis, Canon, Nike, Toyota … | CRO agency owner + scaling €90k MRR subscription business | 🎓 PhD dropout | @koganpage book author

◷ Joined July 2008▧ 396 Photos and videos
Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·24 Jul 2025

CRO ideation workshop at Canon, London How to generate new, meaningful ideas? - Speak to real human customers - Gather real humans from different parts of biz - Outline clear opportunity, based on what real humans are trying to get done Ban any "use AI" easy cop out ideas.

Post attachmentPost attachmentPost attachment
Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Nov 4

I consider myself a half decent conversion copywriter, but when I’m stuck I ask my AI: “What would Joanna Wiebe do?” Sales copywriting OG.

I'm 47. I'm an 8-figure copywriter. I owe it all to the world's most boring strategy. Here's what I do (& how you can too):

Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Oct 26

“Most CRO test screenshots are bullshit” 🎯 @andrewjfaris great pod w/ Mars Men head of ecom He lists the most common issues, but this quote nails it: “Doing CRO is easy, doing CRO the right way is hard”

Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Jul 19

Sol Medium -> sweet spot intelligence vs cost per task Fable -> if you want absolute highest intelligence, cost secondary Sol High -> near-top intelligence at far better value Sonnet 5 -> meh Opus 4.8 -> overpriced vs Kimi K3, basically an expensive Sol High 👇

Post attachment
Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Jun 3

Remember this the next time you ask vanilla chatbot to summarize customer reviews Image stolen off Reddit

Post attachment
Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Apr 8

One of my first CRO clients was 16 years old at the time. That was 15 years ago. He did well, could’ve retired by now but still grinding nights and weekends. I love that founder energy.

Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Apr 6

Why design tasks burn so many Claude tokens and what to do about it. “make it bigger” / “center it” etc reload the entire convo Every mcp loads its definitions in every message, even if not used Select only the specific Figma component and other tips yummy-design-sprint.notion.s…

Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Apr 5

Tip: use Gemini CLI with @obsdmd instead of Claude Claude + Obsidian having a moment on X … but Claude tokens are precious You get ample free requests in Gemini CLI, if you want 3.1 pro it’s pennies via the api.

Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Apr 4

Bros posting all the Gemma 4 hype, have you actually used it?! Been playing with it for 2 solid days, seriously underwhelming vs the hype.

Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Mar 30

Bro we can see AI wrote your tweet Why TF would I bother reading it if you can’t be bothered writing it? I speak to AI all day, I come here for human voice

Johann Van Tonder avatar
Johann Van Tonder@chat_JVT·Mar 29

What I’m trying to figure out: where do open (read cheap!) models fit in? - GLM-5 for planning, refactoring - MiMo-V2-Flash, Devstral 2 highly praised for coding and popular on Openrouter You don’t need frontier models for everything.

I’ve used Opus 4.6 and GPT 5.4 on a mix of projects since release, and want to break down where I think they uniquely excel. It’s more nuanced than you’d think! Rigor of code - GPT 5.4. It goes the distance validating its work without asking. Opus needs explicit instruction to do this, and even then, it misses more edge cases. Clarity of code - Opus 4.6. Claude is a better communicator, which carries into the code. Variable names are clearer and less mechanical, which improves reviewability. This is very important since code review is the bottleneck for most engineering teams. It also adds the right amount of doc comments. GPT simply never comments or explains its work; it’s like working with an obtuse engineer that wants the solution to speak for itself. Sometimes it does, other times not. Similarly, rigor of plans goes to GPT 5.4, while clarity of plans goes to Opus 4.6. An interesting point though: GPT performs better talking through a strategy without a plan, while Opus needs planning mode to put in any rigor. I find myself forgetting plan mode altogether using GPT 5.4. Quality of research - toss-up. Opus spends longer researching with web search, but GPT spends longer studying the existing codebase. You may think codebase research matters more, but researching how others solve the same problem can be just as important. Maybe more important for greenfield. Quality of conversation - Opus 4.6. It’s just better to talk to, which matters using these things everyday. GPT 5.4 was clearly trained to challenge the user more, which results in a tendency to *always* say you are wrong. I’ve had bizarre interactions where GPT claims something is “not quite right,” the restates exactly what we’ve decided on in the last turn. On a personal level, it’s annoying. On a practical level, it makes iteration on a plan slower. THAT SAID, it takes sufficient pushing for Opus to challenge your thinking in this way. Simply say “I’m impartial” and ask questions to avoid that, as you would a person. Overall winner - Opus to make it work, GPT to make it good. I don’t have a good system of when to switch tools, but on average, I prefer Opus early on and GPT for optimization and discussing architectural decisions. Opus is also better for any design related tasks (but state management in frontend apps is better handled by GPT).