First movers have a huge advantage over those who follow them, but this advantage is no guarantee of uninterrupted business success. Technological dominance can quickly melt away, strategic short-sightedness punishes you, significant brand recognition does not protect you from a better product either. In the case of Novo Nordisk, we were recently able to see how a “first-mover” company loses the competition to its successors. And OpenAI can learn a lot from the competition against Anthropic about what mistakes it should not make.
At the cost of huge resources and many years of work, a huge technological innovation is realized, and a company breaks into the market with it ahead of everyone else. The product is a bombshell, millions start using it immediately, which means huge profits for the pioneering company. The new technology becomes one with the brand, becoming the hottest, newest part of American pop culture. Although the initial success is great, the potential reach means a growing market for decades to come, which is why the capital markets expect a huge surge from the results, valuing the company at nearly 800 billion dollars. At first, the company appears to be untouchable and has a bright future ahead of it, maintaining its tremendous reach through better technology.
Then the others come.
The above description could also describe the situation of Novo Nordisk and OpenAI at different times. Novo Nordisk Since its 2024 peak lost 75 percent of its market value and despite its pioneering role already not the biggest actor in the market for GLP-1 drugs. OpenAI also pioneered the market for large language models with ChatGPT, and soon hundreds of millions of people were using their product, for many the word AI became synonymous with ChatGPT. Still, the company’s leadership role in the world of AI is increasingly being questioned, with the development of competitors such as Google, Anthropic and Meta.
This article investigates what mistakes the Danish company made that lost its leading position, and how OpenAI can avoid them, i.e. what can OpenAI learn from Novo Nordisk?
The pioneering role
The companies that are the first to burst in with new technology and create a market advantage are called “first-mover”, which sounds best in Hungarian as “pioneer”. In theory, we can identify four decisive advantages (“first-mover advantage”) from the pioneering role.
- The technological advantage knowledge acquired by can maintain the technological advantage due to the learning curve, thus helping to maintain the pioneer’s advantage. In the pharmaceutical industry, patent law provides protection for the commercialization of innovation, but for a rapidly developing technology, this does not provide sufficient protection, as we will see with Novo.
- A pioneer understands better than its competitors a market’s defining key resources – production capacities, raw materials, AI researchers, GPUs – so you can consolidate your advantage by grabbing them.
- The product that reaches users first can build brand loyaltyif the product’s quality and pricing are so satisfactory that users have no need to switch, and thus stick with the first one they tried.
- A large market share can reduce the cost per productwhich helps expand innovation. This is an important competitive advantage especially for smaller companies. At the same time, pharmaceutical and tech companies with huge capital can operate at a loss for years, so pioneers do not gain a significant funding advantage, which makes this factor less important in these two markets.
On the other hand, the general advantages of the following competitors can also be identified, which can become the pitfalls of the pioneer. They are the first to enter the market
they bear the development and market building costs and risks – they are the ones who test the appropriate product model and educate the market, while the followers can enter the already validated market with a cost advantage already aware of this. The most common pitfall is when the pioneer’s management becomes rigid, clings to the original innovation, and lags behind when the technology takes a new direction.
The similarities
At first, the parallel between two very different industries may seem strange, but the established competitive situation and market dynamics show a rather similar picture of the situation of Novo Nordisk a few years ago and the current situation of OpenAI. Both companies dominated more than half of a newly formed market in the first years, Ozempic and ChatGPT became ingrained in pop culture and became synonymous with the concepts of consumer medicine and AI in public discourse.
However, in a rapidly expanding market, the users can find insignificant differences between the competitors’ products and the “original”, and it would be difficult to create a real switching cost in the B2C market due to the characteristics of the products. No decisive cost or funding advantage emerged from the pioneering position, as Eli Lilly in the pharmaceutical market and tech giants in the AI sector are willing to devote at least the same, if not more, resources to the competition.
In addition, the literature differentiates industries based on the speed of market growth and technological development, and the extent of the pioneers’ advantage: the pharmaceutical industry and artificial intelligence belong to the extremely fast category in both indicators, where it is the most difficult to maintain the leading position.
What did Novo Nordisk screw up?
With the explosion of Ozempic, Novo Nordisk appeared as the only player in the market of consumer medicines, but even in 2024, with the entry of Eli Lilly, it held a market share of around 60%. We are not talking about a small company that did not have the organizational structure behind its sudden success, but about a nearly 100-year-old giant in the pharmaceutical industry – yet in 2 years Novo’s products only reach a third of the market, Eli Lilly is now clearly the leading GLP-1 manufacturer.
Novo Nordisk’s biggest mistake was the severe under-planning of scalability and manufacturing infrastructure, leading to a global supply crisis by 2024 and a de facto loss of leadership.
As demand grew exponentially, the company was unable to secure the sterile fill-seal capacity that proved to be a key resource, creating a critical market gap. This gap was exploited not only by the powerful Eli Lilly, but also by the mass appearance of compounded (quasi-biologically copied) products: as Wegovy and Ozempic became in short supply, the regulators temporarily allowed the “copying” of the active ingredient despite the patent, with this Novo lost control over its own market, as many cheap alternative products were launched.
This logistical blunder was aggravated by mistaking the direction of technological development:
the management had too much faith in the dominance of semaglutide, and instead of devoting its resources to the development of the dual (GIP/GLP-1) active ingredients in time, it focused on perfecting the tablet formulation (Rybelsus) as a priority.
During the Ozempic hype, Novo Nordisk’s stock traded at a P/E ratio of around 30-40, so the Danish company could have taken advantage of the high valuation to diversify itself with acquisition activity. By bringing Ozempic to the market, Novo showed surprising courage, as it threatened the market for its own diabetes drugs, however, in the case of semaglutide, they stuck with the research direction due to the risk of cannibalizing their flagship product. Thus, with tirzepatid, Lilly was ahead of them not only in quantity, but also in quality – in comparative studies, Lilly’s products usually showed greater weight loss, so doctors in contact with consumers and informed by efficiency data turned to the competitor’s product.
Finally, the weakness of the regulatory and communication strategy: Novo could not effectively deal with the “Ozempic stigma” (the consumer drug is the easy way out for the lazy and wealthy people to take over diet and exercise) and the pricing controversy, which created political and insurance resistance – Novo lost the advantage of brand awareness due to public discourse.
Meanwhile, Eli Lilly has already arrived on the more educated market with a cleaner, better positioned model and the more efficient product, Mounjaro.
What OpenAI can learn from this
Based on Novo’s technological insistence, if OpenAI only focuses on scaling transformer models (more GPUs, more data) in 2026, while a competitor (e.g. Anthropic or Google) introduces a fundamentally new algorithm or a new kind of user experience, then OpenAI will meet Novo’s fate: they will have the best-known product, but not the most efficient one. Book by Karen Hao based on this, the dominance of GPT models and development through scaling was not at all a clear outcome of AI research. If the technology takes a new direction thanks to a breakthrough, OpenAI should be ready to cannibalize its flagship product, ChatGPT.
The WSJ in his April article we read that investors and the CFO expressed concern about the scale of OpenAI’s investments, while the number of users and revenues of the company do not reach the set goals. However, from a strategic point of view, I think it is essential to tie up these key resources, so OpenAI rightly insists on ordering chips, avoiding Novo’s number one stumbling block, the lack of building infrastructure to match the exploding demand. The company has already lost the other key resource it initially had – the know-how accumulated by the original research team could have been a long-term competitive advantage if they were able to retain the resources.
The mistake turned out to be critical, since the key people of the biggest competitors Anthropic, Meta, or Google Deepmind are former OpenAI employees.
A seemingly softer but still important conclusion is the importance of managing political relations and public opinion. Novo lost its initial hype when negative side effects, high prices, and the use of Hollywood celebrities brought more and more criticism to the company’s products. In a similar way, OpenAI is starting to become a close enemy, it has lost the image of the original non-profit, helping humanity, publicly operating research company due to Sam Altman’s arrogant statements, state ties, the company’s nebulous financial background, and the negative perception of the Elon Musk trial. Meanwhile, Anthropic immediately came forward in the media as a security-focused, user-friendly and friendly AI company.
Of course, the balancing act is not easy: for the time being, all big-tech and AI companies won the support of the White House by promising huge investments and emphasizing the competition against China, while Anthropic temporarily lost its political position in the His conflict with the Pentagon because of
OpenAI is still leading, but it is already feeling the danger
Although the popularity of Gemini at the end of last year and the advance of Anthropic B2B show the deterioration of OpenAI’s competitive position, ChatGPT is still the number one AI model. The market for solutions offered to companies has not reached the level of consolidation, so the market leader can still take advantage of its brand awareness and create an ecosystem that creates significant switching costs.
Sam Altman in December announced a “code red”. due to Gemini’s B2C breakthrough, focusing on improving ChatGPT’s user experience. Due to the rise of Anthropic B2B, it needs to turn to agentic AI solutions specialized for the OpenAI industry in a state of full alert in order not to lose the competition in the corporate sphere – we are already seeing signs of this with the Codex coding agent.
Keeping the B2B market share is also a more difficult task in theory, since corporate purchases involve more considerations and data-based decisions, so the advantage of being a pioneer is weaker, and for the time being, the monetization of the models is stronger in the segment, so extra resources must be devoted to this sphere.
Novo Nordisk’s history clearly shows that pioneering alone is not enough: lasting success requires constant adaptation, aggressive capacity building and timely recognition of technological trends. OpenAI is still in its early days, but the same risks are already emerging: eroding knowledge advantage, increasing competition and emerging reputational challenges. So the task before OpenAI is clear: can it move in time and even innovate against itself, before others do it for it.
Roland Cossackthe writings of the HOLD analyst you can browse here on HOLDBLOG.
JOGI STATEMENT
The contents of the document do not constitute an investment offer, invitation to tender, investment advice or tax advice, investment analysis, no claim can be asserted against HOLD Alapkezelő Zrt. based on its contents, and HOLD Alapkezelő Zrt. assumes no responsibility for them.
