- Startup Spells 🪄
- Posts
- OpenAI's 12 Days Of Launch: Christmas Marketing Masterclass
OpenAI's 12 Days Of Launch: Christmas Marketing Masterclass
PLUS: Meta Prompting with o1, o1 Pro Mode, and ChatGPT Pro
OpenAI's 12 Days Of Launch: Christmas Marketing Masterclass
Many new businesses dream of a big launch. They work hard to make a perfect product. Then, they show it to everyone and hope for the best.
This is risky. A better way is to launch repeatedly.
OpenAI recently did this with their "12 Days of Ship-mas." They announced something new each day.
On Day 1, they announced ChatGPT Pro for $200 per month. This gets you unlimited o1 Pro requests and unlimited Advanced Voice.
OpenAI Day 1 - ChatGPT Pro
On Day 2, they announced Reinforcement Fine-Tuning.
On Day 3, they announced Sora. Sora can create videos, generate content from text, or enhance your own videos.
OpenAI Day 3 - Sora
Why Big Launches Often Fail
A perfect launch sounds great. But, it causes 2 big problems:
Constant Delaying: You try to make everything perfect. You never start. You add features and change things. You think it's not ready. This wastes time.
High Risk: You put all your hope in one try. If the launch goes poorly, you waste time and money. You have little to show for it. Many founders spend 6 months to 4 years making their product. Then, they have a bad launch, lose motivation, and quit. That's what can happen if you only launch once.
Launch Many Times Instead
Don't think of your launch as one big day. Make it part of your business. Release early versions. Get feedback. Make changes. Then, launch again. Talk to your users. Don't just make announcements.
This works because you get to:
Practice Your Message: Each time you launch, you practice your pitch. You learn what works. You fix what doesn't.
Improve Your Product: Early launches give you feedback from real people. This tells you what's good, what's bad, and what to fix. You build a product people need, not just what you think they want. You could release a simple version, an MVP, to collect feedback.
Find Your Customers: Launch to different groups. You might find that the wrong people like your product. Find the right ones.
Different Ways to Launch
There are many ways to launch, according to Kat Mañalac.
The Simple, Quiet Launch
This gets your product out there simply, even if it's not done.
What You Need: You don't need much. Just a basic website. Put your business name on it. Add a short description. Add a way for people to contact you and a call to action.
Example: A startup named Zinc has a simple landing page. It has a "Get in touch" button. They can collect leads and see if people are interested.
Tools: Product Hunt's Ship lets you make pre-launch pages. You can build an email list.
Example: Designer School uses Product Hunt. Their page has a short description. It has a "Subscribe" button. They build interest and collect emails before they launch.
The Friends and Family Test
This is your first chance to get feedback.
Get Early Feedback: Show your Minimum Viable Product (MVP) to people you know. Watch them use it. Ask questions.
Example: Reddit first launched only to people in their Y Combinator group. The Wayback Machine shows how basic it was. This small launch gave them early feedback.
Note: Friends and family might be kind. They might not be your real customers. But, it's a start.
Test With Strangers
Now, get feedback from people who might actually buy your product.
Get Real Feedback: Make sure your product solves a real problem.
Example: Lugg helps people move. The founders rented a truck. They went to IKEA. They found people struggling to load their cars. They pitched their service. If the customer said yes, they used their app to "help" with the move. (The app wasn't real yet.) This gave them great feedback.
Use Online Groups
Online groups are full of people who like new things.
Get Specific Feedback: Use groups related to what you're building. Get feedback from interested people.
Example: Y Combinator companies often launch on Bookface first. This is their private network. Then, they use the Startup School Forum. It has 40,000 people.
Success Story: A friend shared Magic on Reddit and Hacker News. 40,000 people signed up in one weekend. This shows the power of online groups.
Data: Hacker News, X, and Reddit are now as good as big websites like TechCrunch for getting users.
Rules: Learn the rules of each group. Join discussions. Be open to feedback.
Example: One founder wanted feedback from women. He asked someone in Alpha, a group for women in tech, to post for him. This shows the power of using groups to reach specific people.
Writing Tips: Write like you talk. Don't use fancy words. Be real. Share interesting stories. People on Hacker News like learning new things.
Example: Share unusual things you learn from users. Did people use your product in a surprising way? Share it. Explain what you did about it.
Ask Questions: Say what you're an expert in. Ask people to ask you questions about it. This helps guide the conversation.
The Waitlist Launch
This makes people excited about your product.
Create Excitement: Make a waitlist. Let people move up the list if they do certain things. This makes people feel like they're part of something special.
Example #1: Magic had a waitlist. People who tweeted about it moved up the list. This helped them get more attention.
Example #2: Superhuman is a popular email tool. Current users can refer new users. This lets new users skip the waitlist. Their emails say "Sent by Superhuman." This gets them new users.
Use the reach of others to get attention.
Work With Influencers: Find bloggers, journalists, and people on social media who are popular in your field. Offer them early access to your product.
Example: Joy makes wedding websites. They found top wedding blogs on Google. They emailed the bloggers. They asked to be included in "best of" lists. This got them many early users.
Avoid Paying: Some bloggers want money. You don't need to pay. Offer them special content or early access. Build real relationships.
Pre-Orders
This is good for physical products.
See if People Want It: Use Kickstarter or Indiegogo. See if people will pay for your product before you make it.
Example: Sheertex makes strong pantyhose. They made a video. They pitched it to the press. They launched on Hacker News. They told everyone they knew. This helped them get a lot of attention and money.
Keep Launching New Things
Keep your users interested. Release new features or products often.
Keep People Coming Back: Don't just launch once. Keep making things better. Release updates. This keeps people excited.
Example: Stripe is great at this. They launch new things all the time. They talk to the community about it. They announced Stripe Atlas on Hacker News. They joined the discussion. They shared their thoughts on their blog and social media.
Example: Glossier is a beauty company. They have a plan for launching new products. They use every channel (communities, social media, press, ads) for each launch. This keeps people talking about them.
Build Your Group
A strong group of users is important.
Build an Email List: Start from day one. Ask everyone if you can add them to your email list. Send them updates.
Example: Gaddy Avron from Symetra had a big email list. He added everyone he talked to about his business. When he was featured in TechCrunch, he emailed his list. He asked them to share it. Even people who didn't invest in his company shared it. This shows the power of an email list.
To Summarize
Launch is a process, not one day. Don't aim for one perfect launch. Launch many times.
Don't get stuck on one big launch. Start small. Get feedback. Improve. Launch again.
Use all the ways to launch. There are many ways to get your product seen.
By launching many times, you can keep yourself in the conversation. When people need to do a particular action, they associate it with your product. For example, search is attached to Google. Rides are attached to Uber.
OpenAI's 12 Days of Launch is clever because it takes advantage of the days before a big holiday. Christmas is celebrated widely. Launching every day for 12 days is a brilliant marketing move. Instead of one big launch, they spread it out over 12 days. This makes it a spectacle.
Top Tweets of the day
1/
Need a Chief Memelord?
I'm running a Cyber Meme Monday only for startups.
My team of cracked memelords (AKA the Memelord Mafia) will deliver you:
• 1 meme/day for $420.69/month (makememes.money/420)
• 3 memes/day for $699/month (makememes.money/699)Why should you… x.com/i/web/status/1…
— Jason Levin (@iamjasonlevin)
5:03 PM • Dec 2, 2024
The advantage of building in public like this is once you start an agency, you already have built in trust from your audience.
He built an audience by posting memes. Now he's starting a meme service. People trust him. They've seen his work. He's charging $420.69 a month for 30 memes. Or $699 a month for 90 memes.
Some people will pick the cheaper one but the more dollar per value is one the higher tier. He set it up that way on purpose. It's the Decoy Effect in action.
Not too bad for an hour of work. Already made $5k/month for doing memes as a service.
He could even make it a marketplace where people sell memes. Like Fiverr, but just for memes.
2/
New ideas only come from fringe people. But fringe people are never fringe in just one dimension. In psychometric terms, 100% trait Openness.
— Marc Andreessen 🇺🇸 (@pmarca)
6:02 PM • Nov 17, 2024
True. Warren Buffet might seem like a saint with his talks but he'll eat your business for breakfast if it came to that.
3/
I’ve noticed a lot of business lessons I need to relearn at each level.
How I started:
Focus on one thing
Hire smart people
Think long termEach of these lessons I’ve had to relearn as companies scaled because:
Focus is even harder when opportunities grow like weeds. They… x.com/i/web/status/1…
— Alex Hormozi (@AlexHormozi)
5:51 PM • Nov 1, 2024
One thing I love about Alex Hormozi is he repeats the same lessons over & over again.
I've learned if someone keeps repeating something, internalize it so you don't have to make the same mistakes.
If it is important for someone as successful as him, it is probably a great advice.
Rabbit Holes
Meta Prompting with o1, o1 Pro Mode, and ChatGPT Pro (Compute on Compute) by IndyDevDan
Search Engine Deoptimization by Edward Sturm
Warren Buffett's Efficient Networking Filter by MFM Vault
What’d ya think of today’s newsletter? Hit ‘reply’ and let me know.
First time? Subscribe.
Follow me on X.
Reply