THE INFINITE FOCUS GROUP: AI IN MARKETING
Hold onto your artisanal hats, friends.
Because your marketing life is about to change, drastically. Like that time you tried to cut your own bangs. But in a good way.
In B-school we learn that marketing is everything. The problem? We can't read minds. It's a bummer.
But what if you COULD know What People Wanted? Like Mel Gibson in that movie which shall not be named because, you know, Mel Gibson issues.
Enter: The Infinite Focus Group™ (Trademark pending. Kind of.)
Here's something people don't know about ChatGPT:
It's an absolute wizard at imitating humans. Seriously, it's almost impossible to tell if it's a bot or your Aunt Karen writing that scathing Yelp review about the lack of gluten-free options at Olive Garden.
You can tell ChatGPT to conjure up virtually any demographic. Any language. It's like having a genie, but instead of wishes you get frighteningly accurate consumer insights.
Here's how it works:
step 1: set the stage
Let's say you're selling Conor's Organic Banana Yogurt (yum!). Ask ChatGPT to magic up a diverse group of 5 people in a grocery store in Kansas City.
Bam. You've got Ethan the 27-year-old PT, Camila the mom of three girls under 7, Phil the retired plumber...you get the idea.
Step 2: Focus Group Frenzy
Treat them like a focus group. Ask why they would or wouldn't choose your product. The human-like answers will blow your mind.
And we're just getting started.
Do this a hundred times if you want.
Have ChatGPT record the data in a spreadsheet.
Analyze it for insights and recommendations.
It's like having a team of AI marketing ninjas.
But wait, there's more.
Want to go deeper?
Take one of your focus group folks, or create your ideal target audience. Let's say working moms with young kids. Have ChatGPT whip up an Isabelle.
step 3: truth serum time
Now here's where it gets juicy: inject Isabelle with truth serum. Humanely, of course. We're not monsters.
Ask Isabelle what she REALLY thinks about your product.
Why she makes the choices she does. Is it to impress colleagues? Parental shame from a snide comment?
This is marketing gold. You’ll see immediately how ChatGPT goes deeper into the psyche of how we think, how we honestly make decisions.
step 4: fridge raider
Want more gold?
Respectfully pop over to Isabelle's house. Take a gander in her fridge.
Have ChatGPT analyze what those other choices might mean for your yogurt's appeal.
Zoom out. Check a hundred more fridges. Gather that sweet, sweet data.
I know what you're thinking. "But Conor, is it as good as a REAL focus group?"
Friends, test it yourself. ChatGPT focus groups are more honest (with the right prompts), can be any exact demographic, language, city, or product. You set the rules.
If you can’t tell it from a human response…then yes. It’s as good as a human response.
And the cherry on top? You can feed all that delicious data right back into ChatGPT for analysis.
Mic. Drop. (Carefully though, those things are crazy expensive.)
So go forth and focus group to your heart's content. Responsibly, of course. And also try some of that banana yogurt. Lemme know what you think.
Till next time, magicians!
AI NEWS OF THE WEEK
1. Amazon and Anthropic Go Big
Amazon has invested another $2.75 billion in Anthropic, this is after their $1.25 billion investment back in September. They’re counting on organizations accessing Anthropic a ton via their cloud, Bedrock. Let’s see if this pays off for them. That’s a lotta scratch.
2. HeyGen Keeps It Real
The phenomenal video generator HeyGen, which creates hyperrealistic video/audio of you after only a few minutes of sample, is actually making money. According to The Information, they’re making $20m in annualized revenue and are getting $60m at a $440m valuation. That’s crazy impressive compared to their competition.
3. Elon Groks the Masses
Grok will be available to mere Premium subscribers on X! I’m excited about this - I was paying the extra cash but it wasn’t worth it (sorry to editorialize). More folks will be able to see what Elon’s building in the LLM space.
Generative AI Tips
I love ChatGPT. I do. But it can get…wordy.
My buddy Justin Fineberg posted about this little hack which has become a favorite.
When ChatGPT starts giving you long answers and you just can’t take it - add these two simple words to your prompt:
No yappin’.
Tight, concise answers.
You gotta love it.
That's all for now, friends! See you next time.
