👋 Hi, it’s Naiomi and Erik. Welcome to Demand our weekly newsletter to help you become a top 1% marketer. In today's article, we are going to cover OpenAI’s meteoric rise and the GTM motions which fueled it.

This is the final post in the four-part series on product launches. The feedback from you was clear following the Airbnb deep-dive: let’s see an example of a company in the trenches figuring out their GTM and launch strategy. Who is doing hand to hand combat that we can learn from? We heard you.

What more relevant company than OpenAI to dive into.

Part 1 | PLAN: Start with the end in mind 🧠

• Foundational go to market frameworks underpinning product launches

Part 2 | EVOLVE: How launch strategies evolve as companies grow 🌳

• We’ll cover how launches differ by company stage including investment level 

Part 3 | DEEP DIVE: A behind the scenes look at how Airbnb launches products 🚀

• A case study on how Apple inspired Airbnb’s twice-a-year releases 

Part 4 | DEEP DIVE: OpenAIs Ascension: From Serving Scientists to 180 Million Consumers (☝️ this post)

• Decoding OpenAI’s product strategy & GTM motions to reach mainstream

A Meteoric Rise

OpenAI is one of the most consequential companies in recent history. Its AI chatbot, ChatGPT, scaled to 100 million users in 2 months. They’re threatening the greatest business model of all time. And there are at least two Hollywood movies to be made – one from the failed governance which led to a coup last November to one about the founding saga between Elon, Sam, Sergey, and IIlya. 

How are they so successful in spite of all the noise? Putting aside perspectives on politics, the FUD that has been unleashed on AI, and the for-profit vs non-profit fiasco, we want to focus on why they have been so successful and how marketing and GTM fueled this success. 

In this article we’ll cover the following: 

• 🚀 What products have OpenAI launched and how did they launch them?

  • How they evolved from largely a technical research institution to a commercially-focused modern enterprise

  • How their business model has evolved as product and target audience evolved 

• 🏆 Why have they been so successful? What are the conditions, frameworks, and tactics used by OpenAI?

  • How their team and culture played an important role in setting the right conditions for success

  • How their strategy evolved from serving a highly technical niche audience to a broader market across consumers and enterprises

  • The role that marketing and public relations played in driving growth 

Crossing the Chasm from Researchers to Consumers 

Since its founding in 2015, OpenAI has built two types of products: research & development and commercial. The foundational research and development products support its core mission of developing artificial general intelligence benefits all of humanity which underpin the commercial and consumer products we know today. 

OpenAI has evolved from research lab to AI frontrunner amongst the tech giants. Initially focused on academic breakthroughs with no intent on making a profit, they've exploded in popularity with consumer products, becoming the fastest product to reach 100 million users in just 2 months. With this growth, their target market has expanded from research scientists to enterprises and consumers.

Fueling Enterprise Growth

While its consumer growth is impressive, OpenAI’s enterprise adoption has exploded. In just four months of launching ChatGPT for enterprise, they reported 260 customers totalling 150,000 enterprise users. Altman announced they have 92% of fortune 500 companies using OpenAI products. To capture value they’ve rolled out their business model, which by December 2023 crossed $2 billion. 

That is a ~900% increase YoY in revenue. Where is this revenue coming from? Revenue streams include: 

• API Access: OpenAI licenses their AI technology, like GPT-3, to companies for various applications. Examples include Asana, Bill and Melinda Gates Foundation, Klarna, Lowe’s, PWC, JetBlue, and Amgen. 

• B2B Subscriptions: they have two segments: Team for smaller companies and Enterprise for larger companies. For ChatGPT Team they charge $25/user/month billed annually and for ChatGPT Enterprise the price is not listed on their website, but quotes shared online range from $20 to $60/user/month with some claiming on Reddit a minimum of 150 seats per company. 

• Consumer Subscriptions: such as ChatGPT Plus for $20/month. 

• Grants: from government agencies, philanthropic organizations, and other institutions that support their mission of developing safe and beneficial AGI.

• Strategic Partnerships: similar to API Access, but more hands-on and custom co-development with a partner such as integrating OpenAI’s technology into Microsoft’s Bing product. 

Key Ingredients of the GTM Rhythm 

So how has a non-profit with no interest in commercial success evolved into one of the fastest growing startups catapulting AI to the masses and an impressive valuation of 80 billion? In addition to getting the right governance model in place and developing consumer and enterprise grade products, it's been a combination of factors including team, culture, and building their GTM muscle.

1. Talent Density

The culture and talent density of OpenAI is unique, even by Silicon Valley standards Back in 2009 Paul Graham penned an essay about 5 most interesting startup founders of the last 30 years which included the usual suspects like Steve Jobs and Larry and Sergey, but it also included a 24-year Sam Altman who was the founder of Loopt.

The team also features heavy hitters such as Ilya Sutskever who has worked in the labs of Geoffrey Hinton and Andrew Ng while a student and was previously a researcher at Google Brain. Sutskever was so important to Google, that when Musk poached him to join OpenAI, it led to Musk and Page no longer talking or being friends as reported by biographer Walter Isaacson. For marketing, this talent density has led to the hiring of Krithika Muthukumar last August, who comes by way of Stripe and Retool. They are  continuing to hire and develop their go to market capabilities, as highlighted by the 24 open roles on their Go To Market team.  

2. Grand Ambitions 

The team has stated they want to “Solve Impossible Problems”. It’s a modern day Manhattan Project with Altman orchestrating the grand strategy:

3. Researchers that build

Researchers aren’t just publishing papers and models into the abyss, they are building core products. The sprint team consists of Researchers, Engineers, Designers, and Product Managers (aka DERP). This has enabled them to ship products incredibly fast. 

4. Small Teams

They have small, focused teams with clear constraints. Altman believes large teams suffer from dilution of focus, internal politics and slow decision-making. Altman recently remarked:  

Compare the roughly 1,000 employees at OpenAI to Google, which had nearly 200,000 employees in 2022. 

5. Low Ego 

The team prioritizes output vs personal agendas. As Peter Yang noted, almost everyone at OpenAI has the title: “Member of the technical staff.”

OpenAI GTM Ingredients

Just 4 years ago OpenAI’s GTM cadence was largely absent. In 2019 they announced a robot that could solve a Rubik's cube by hand, which is an incredible engineering feat, but it garnered a meager 10,000 likes on Twitter with engagement mostly limited to the science and research communities.

Fast forward to today, they’ve built a GTM machine. OpenAI’s Sora announcement in February received 142,000 likes and 100 million impressions on Twitter and tens of billions of earned impressions. The market is loving it. Their newly established tenets of GTM include:    

1. Show Don’t Tell

OpenAI embodies this principle to their core. Instead of overly polished launch videos that don’t show the actual product, they put the actual product front and center in their launches. One moment captures this principle flawlessly. Let’s rewind to February 15, 2024. Love was still in the air. More so for OpenAI than Google. Around 9am, Sundar Pichai, CEO of Alphabet and Google announced Gemini 1.5 Pro with the following tweet: 

128K-token context window? Next-gen model? Mixture-of-Experts (MoE) approach? Efficient training & higher-quality responses? 

Yawn. That might be the tastiest feature soup ever made. But it didn’t land in the market. The CEO of a nearly $2 trillion market cap company could only generate a paltry 825K views.

Four hours later Sam and OpenAI announced their text-to-video product, Sora. No specs. Just seeing the actual product. The first prompt was “Beautiful, snowy Tokyo city is bustling..” which produced: 

Beautiful. But what happened next was Jobsian. 

Mr Beast and other influencers replied to Sam encouraging new prompts. Within an hour, Sam would reply with incredible clips generated from Sora.

Such as… 

…which led to:

and this…

…which led to:

You get the point.

In total, they published 48 clips which generated nearly 150M impressions in 24 hours. These demos captured the imagination of early adopters and press. Seeing Sam reply real time to showcase the product was phenomenal.It generated significantly more interest, and by extension credibility in OpenAI.

Further, they haven’t over-extended in their demos. They’ve avoided huge PR nightmares, unlike Gemini from Google, because they have the right guardrails in place. The initial ChatGPT model launched in November 2022 was text-only and its training data was only up to September 2021. They made these constraints known to users. Even with this disclaimer and warnings about hallucinations, over 1 million people used it in the first 5 days.

2. Make it human

They have taken a product highly technical and often marred with skepticism and fear, AI, and made it hyper accessible to everyday consumers. With the launch of ChatGPT in November 2022, they made AI approachable. It’s so easy your grandmother could use it. You didn’t need special access. You didn’t have to join a Discord server. Or need an API key. Or join an extended waitlist or enterprise account. The ChatGPT launch demonstrated the make it human principle. 

Most notably: 

• Ease to onboard: it was free to use, you only needed an email address and password. 

• Ease of use: With the initial ChatGPT launch in November 2022, anyone with access to a browser could try the product. The UI was clean and used a form factor people were familiar with – search. They also kept the interaction simple – it was strictly textual interaction, so they reduced edge cases and avoided local maximums. 

• A dose of magic: ChatGPT wasn’t just a one-cycle interaction like search which provides results. Unlike search, ChatGPT held conversations, remembering past prompts for a truly magical experience. In the months that followed there were countless examples of folks sharing their prompts on Twitter/X and LinkedIn and posting their videos on TikTok and YouTube. Such as: 

@misscarolineflett

🪄The magic prompt for Chat GPT content creation… If you're looking for ways to get AI and chat GPT to plan your content for you, this vide... See more

There were many more. The term ‘prompt’ and ‘ChatGPT prompt’ spawned hundreds of TikTok videos, thousands of articles and newsletters, and many engaging posts such as this one on Reddit. The fact that this interaction was so shareable and replicable fueled consumer adoption. They were able to bring AI mainstream because moms, kids, and grandparents could participate in this new thing they had only heard about. They took something that was cloaked in mystery and made it accessible and useful. 

In the early months, ChatGPT was often at or over capacity, which restricted access. This scarcity increased interest from consumers and press even further. 

3. Sam as Chief Spokesperson 

Sam is the face of OpenAI and it's a big advantage for them. He's young, ambitious, charismatic and consistently generates press for the company. His appearances at high-profile venues such as major media interviews, Dev Day, podcasts like the Lex Fridman Podcast (twice) and his Congressional testimony have helped solidify OpenAI's reputation at the forefront of AI development.  Sure, he’s not perfect, but he’s been a huge brand lever for the company. 

4. Developers

I mean, obviously. We’re talking about a new computing paradigm. Developers are a P0. From early on OpenAI has been catering to researchers, scientists, and developers. More recently they’ve leaned into Developers in a big way. Which makes sense. OpenAI is building the foundational LLM layer and they want to have an open ecosystem for developers to build custom experiences on top of. It’s what Jaryd Hermann of How They Grow notes as the Layer 2 and Layer 3 companies:

OpenAI is following the developer conference playbook used by companies such as Google (I/O), Amazon (re:Invent), and Square (Unboxed) with their own version – Dev Day. Their inaugural conference was held on November 6, 2023, in San Francisco and featured announcements and demonstrations of feature improvements and new models. They announced GPTs, the ability for developers to create their own custom versions of ChatGPT, and followed that up with a launch of their app store, the GPT Store in early 2024. Their intent is to build a thriving ecosystem on top of their platform to increase value and lock in. And it's working. In January they announced over 3 million custom versions of ChatGPT. 

5. Microsoft Partnership 

Microsoft is entitled to up to 49 percent of the for-profit arm of OpenAI's profits and ChatGPT is a threat to Google’s search dominance, so there are clear incentives for the Redmond-based behemoth to support the upstart. While Microsoft has supported through multiple rounds of investment and research resources, they also have significant relationships with most large enterprises which provide warm introductions for OpenAI. Further, OpenAI can utilize the Azure platform to enable faster distribution. 

Leaps Made, Gaps to Close

While they have made massive strides in their GTM rhythm, they have a number of improvements that they need to make to be world class. These include: 

• Brand: Intentionally build a brand architecture with a clear narrative of where the brand is positioned relative to competitors and importantly, where the brand is going. They can make it even more human by showcasing how ChatGPT or Sora are being used to benefit humanity. Show how it's being used in schools. By small businesses. By creators and artists. By scientists and healthcare workers. 

• Positioning: their positioning is underdeveloped and inconsistent– from their videos, to their website and emails. This is related to brand and brand architecture but ultimately, they will need to create clear, compelling and consistent positioning. 

• Pricing: they are likely leaving money on the table while also preventing adoption with their current price per seat per month and token usage. An emerging trend with AI is pricing per performance. Such as Microsoft’s Copilot for Security which will charge at $4/hour, a pay-as-you-go model which makes it easier to get started. Closer to true performance is Intercom charging $0.99 per resolution for their Fin AI Chatbot.

• Monetization for Developers: to build a true platform and attract and retain the best developers, their ChatGPT store must offer monetization to attract developers. Right now most ‘apps’ are simple reskins on top of ChatGPT. 

• Conferences & Events: the inaugural Dev Day was a step forward, but needs to be more impactful.  Additionally, OpenAI should have a larger presence at industry events. A precursor for the latter will be investing in security & compliance solutions so they can increase adoption in healthcare, finance, and insurance.

What can you take away to be a top 1% marketer?