5 Comments
User's avatar
Ethan Yeo's avatar

Excellent succinct and structured comparison for both models, would be great if this can be expanded into a follow-up series for more systemic levels of understanding why one model is more suited than another for select aspects of marketing. Hopefully making the qualitative conclusions here even more robust to count on.

Expand full comment
John Tay's avatar

That's true. Back to 1st principals thinking - why do certain models perform better with different aspects of marketing. Appreciate this, will have a think about how to structure future tests!

Expand full comment
Vlovsky's avatar

I like how meta this article is.

Again, thanks for the sharing!

Expand full comment
John Tay's avatar

Glad you found it useful! If there's any topic that you want me to explore, let me know too :)

Expand full comment
Nicholas Tolson's avatar

Claude may be technically better on certain things, but I find ChatGPT's features (Custom GPTs, cross-thread context - especially in projects, etc.) make it more useful overall. In the end, I pay for them both, as well as Gemini.

I listened to a podcast with the engineers behind o3. When asked for their tips on using ChatGPT, they said run the same prompt multiple times, because you can get very different results each time. I've started to do this, and it's true. Outside of the scope of this test, of course, but may be something interesting to try if you're dissatisfied with the output of one model or another.

Expand full comment