The TLDR

Highlight: This week, we’re exploring how to use ChatGPT to give yourself a crash course in how awesome brands run website strategy. It’s a transferable technique that you can use to extract best practices from any group of things you want to emulate. Check out our final playbook here.

Musing of the Week: We consider the strained balance between convenience and our right to privacy.

🗒 ChatGPT will teach you to emulate your favorite brands

Over the past few months, we've been hit with a ton of recommendations for all these new AI tools that seem to pop up non-stop. It's exciting from a technological perspective but overwhelming in practice. ChatGPT is still the top player in the game (with over 1.6 billion site visits in June 2023), and its ever-improving capacity makes it the tool we use the most in our workflow, regardless of function, topic, or task.

This week, we’re using ChatGPT and its Code Interpreter plug-in to give us a crash course in website homepage strategy. We wanted it to analyze our favorite examples, recognize nuanced patterns and best practices, and output a playbook in PDF form - and it delivered.

We have experience in consumer brand world and still felt that this AI-generated playbook was novel for a couple of reasons:

• This whole process took about 3 hours. An expert brand copywriter could create a similar (or potentially better) output, but probably in 3-4x the time. The copywriter’s time is better utilized skimming through this competitive intel and translating the learnings for their own work. There’s always an opportunity to dig deeper, but these are great starting points.

• We already have implicit beliefs about the brands we’re looking at, and ChatGPT’s ability to assess them with objective, unfeeling eyes really forces us to look at the data and the patterns across variables.

Try it out for yourself!

🤓 Our Process:

1. Figure out your learning goal and find some examples that are currently executing well.

   • For our process, we wanted to see if there was a pattern to the homepage copy of 10 of our favorite brands (Equinox, Flamingo Estate, Cymbiotika, Chomps, Bigface, Maldon Salt, Architectural Digest, The Ordinary, Nobu Restaurants, Alo Yoga)

2. Export the examples into a file type that can be uploaded and processed by ChatGPT’s Code Interpreter plug-in.

   • We chose to download each brand homepage as an HTML file.

3. Create a prompt (see ours below) telling ChatGPT what you want to learn, context on the examples you’re providing, and the final deliverable.

4. Read through the outputs and refine your follow-up prompts. Coach the program, direct its focus, and ask directly for more advanced pattern recognition.

   • We spent 2-3 hours refining and re-directing until we were happy with the final output and downloaded it as a document.

   • Check out our Modern Brands: Homepage Playbook here.

🥸 Watch-Outs & Notes:

• Make sure you’re using GPT4 with Code Interpreter enabled.

• In one of our earlier prompts, ChatGPT began responding (very confidently) with generic assumptions it was making based on the names of the HTML files we gave it without actually processing the files. (The working dialogue' box will appear when it’s actually opening the files.)

• Remember that ChatGPT is designed to make predictions based on the content it’s seen before. This means that if 9/10 webpage playbooks are generic (which they are), that’s the first thing it will try to replicate. It’s absolutely capable of doing better and deeper analysis, but you will have to prod it a bit in the direction you want.

🧠 Musing of the Week

We're standing at a crossroads, faced with a choice that could define our relationship with technology. The convenience of AI, social media, and targeted shopping all beckon us— sometimes, the way that devices anticipate our needs almost feels like magic. They make our lives more streamlined, from personalized recommendations that save us time to voice-activated assistants that respond to our every command.

Yet, the offer of convenience comes at the cost of our privacy, a fundamental right that we're increasingly urged to relinquish. Our personal data, much of which we don’t even really know about, now fuels the algorithms that power our digital lives. The same devices that make our routines easier also track our habits, preferences, and even our physical whereabouts. As the boundary between public and private blurs, questions arise. How much of our personal domain are we willing to surrender for the sake of convenience?

In this tension between two forces, it's valuable to recognize that the balance isn't fixed. It's a delicate act of negotiation that involves weighing the benefits of effortless living against the potential risks of an exposed existence. AI holds the potential to create conveniences that we can’t even fathom yet, but it is weighed by an equal threat to the privacy we hold dear. Navigating this tension will require mindfulness beyond what was demonstrated during the last wave of technological development.

🙌 If you’re hyped about the generative AI industry specifically, here are some of the coolest roles we’ve seen this week:

• Strategic Account Executive @ People AI

• Account Executive AI/ML @ Flywheel

• Senior Product Manager, Data & AI @ Yohana

• Principal Product Manager, AI @ Asana

• Customer Training & Education Manager @ Seamless AI

🔨 Check out these other AI tools we’ve been playing with this week:

• Scribe - AI that generates process documentation while you work

• Skippet - AI-powered project and data management base for teams

• Ocoya - AI-managed social media content creation, scheduling, and posting

That’s all for this week. See you next Tuesday!

Lorel & Reily