Andy Jassy is driving efficiency within Amazon, which should be facing the innovator’s dilemma; yet, Amazon’s culture, technology, and people (or the lack thereof) will ensure its future.
When Amazon founder and current CEO Jeff Bezos announced that his successor would be Andy Jassy, I admit I thought he had made a mistake. However, Bezos was three years ahead and could see the future. I thought Dave Clarke or Jeff Wilke should have taken over, yet in hindsight, neither was a good fit, and Bezos was hedging the future with someone who built a software business from nothing to billions rather than taking the logistics infrastructure builder or consumer-focused leader that would have been prescient.
The Day Two Version of Amazon
Following the founder of a generational company is hard, I would say borderline impossible. I simply look at what happened at Alibaba, with Daniel Zhang, and how hard it was to follow Jack Ma. We will never know how much input Zhang really had, as Jack Ma was persona non grata with the Chinese government regarding Alipay’s impact on the financial sector in China. Alibaba has a different corporate strategy than Amazon, but the similarities between the situations strike a tone for me. Daniel Zhang stepped down and then abruptly left Alibaba Cloud, and was supposed to manage a fund.
Source: The Verge
Andy Jassy has had to do unpopular things as Amazon was more entrepreneurial under Bezos. The company took many financial positions in the hope of customer growth, but long-term, these decisions needed to change. Jassy likely has had to make tough decisions on partnerships, internal businesses at Amazon that were not making headway against competitors. Jassy grew Amazon Web Services (AWS) at a time when Amazon was more focused on investment in retail and logistics for the future.
I am fully aware that the argument can be made that under Jassy’s leadership, Amazon has become a pay-to-play retail marketplace and that its third-party services costs have significantly increased for third-party sellers. I will counter that with any e-commerce marketplace that is in a maturation phase is likely to be dependent on high-margin businesses such as advertising and third-party fees (commissions and Fulfillment by Amazon (FBA). The narrative that sellers have to pay a 50% tax to sell on Amazon is convenient for regulators and Amazon naysayers.
Amazon is a unique business in that every year is different. Amazon changes yearly to understand the current supply and demand opportunities and to monetize them in a manner that is different than the previous year.
Amazon is a software platform
I have not publicly spoken about this, but in 2024, I changed my perspective on Amazon. I have followed and will continue to follow it, but its e-commerce marketplace does not offer any opportunities for innovation. Over the past five years, the one consistent trend I have observed is that marketplace sellers continue to see their margins erode due to the costs associated with being on the marketplace. I believe it’s fair to say that Amazon’s marketplace business will at continue to grow anything from 5-15%, but that is not interesting,
Amazon’s future is dependent on Amazon Web Services (AWS) becoming a $250 billion to $500 billion business in annual revenue. AWS releases a firehose of announcements weekly and monthly, which I believe is done intentionally to keep the market guessing. AWS has been able to democratize software, unlike any other platform, and continues to ensure that costs are not a barrier to technology development. AWS provides developers with a single platform that can be used to build whatever is needed, whether it’s mission-critical software or sector-specific software.
Amazon invested in its logistics and retail businesses to ensure that it can offer customers convenience. However, that convenience is in low-margin sectors that require significant scale to maintain profitability. Am I the only one who sees that, since Jassy became the CEO and President of Amazon, the focus has shifted to high-margin sectors such as AWS and Amazon advertising? The investment in Prime Video is intended to offer Prime members another value-added benefit, but more importantly, it provides another platform on which Amazon can sell advertising. Amazon is a software platform that provides retail and logistics solutions to sellers and consumers.
Software and, in particular, artificial intelligence (AI) will drive Amazon’s future; nothing else will.
Amazon is Focused, and that Should Scare Competitors
I listen to Amazon earnings calls quarterly. Why? You learn a great deal during the question-and-answer section, when investors and investment banks can ask questions. Jassy is on these calls with Amazon CFO, Brian Olsavsky. Bezos has no intention of ever participating in these earnings calls, and I essentially believe that’s why those earnings presentations look like they were created by a child (they have improved, but it’s still not a presentation that shows a lot of effort).
Under Jassy, Amazon is more focused on profitability than it ever was. Amazon’s retail business is primarily focused on profitability and margin. Amazon has implemented measures such as increasing minimum purchase limits for free shipping, charging for returns, and, more importantly, removing programs that generate a loss. If you listen to first-party vendors who sell directly to Amazon, you will hear that annual vendor negotiations are more arduous now than they were in the past. Brands that sell less than $5 million via 1P are of no interest to Amazon vendors.
When Jassy discusses topics such as cost to serve in the annual shareholder letter, it’s clear that, within Amazon, accountants and finance are more important than they were a decade ago. Amazon is now in a position where it can invest billions into AWS data centers and generate profits every quarter.
AI and the future of Amazon
Amazon, through its AWS business, made some interesting decisions regarding AI. When Jassy was in a photo to announce the significant investment in Anthropic, it was clear to me that this is different. Amazon has decided to act as a marketplace to aggregate various AI models and offer them to developers. I
nvesting billions into Anthropic, Amazon made a decision that it had never made before. Which is another sign of the Jassification of the business. Amazon used to believe that it should build everything itself; however, since Jassy became its CEO and President, Amazon has either partnered with or invested in startups for the future. I believe that this is Amazon’s management team understanding that they lost significant talent to competitors such as Walmart and others but Amazon generates so much cash quarterly, they can use that cash to invest in companies without the need to place risks on their balance sheet.
Investing in Anthropic through AWS credits allows them to utilize their own resources without incurring expenses for hiring or competing with other AI startups. Let me be clear: the current AI race is only going to benefit incumbents who own cloud infrastructure, offer models and develop their own models, and provide applications to third parties. There is not a long line of companies that will benefit from AI.
Amazon is also developing hardware, such as graphic processing units (GPUs), that are required to do inference and training of data. NVIDIA, whilst an Amazon partner currently, needs to understand that Amazon will always want to democratize access to technology.
In the latest funding round for Anthropic, Amazon utilized its hardware and GPUs in the transaction. Yes, I know not many are using Amazon GPUs, but Amazon will want to hedge itself from a future on which they are dependent on a third party (sounds familiar). I believe that in the next decade, Amazon will either acquire or invest in companies that build GPUs to negate the dependence on Chinese-made technology. The geopolitical headwinds are not spoken about enough in the AI race.
Jassy has been clear on earnings calls over the last two years that AI is very important for Amazon’s future. In recent calls, he has even said that the process of getting customers to AI is dependent on them first moving to the cloud. These enterprise customers take months to years to move to the cloud and then require hand-holding to start using AI developed in-house.
Amazon has clearly made a decision to utilize AI in a B2B manner, rather than a B2C manner, unlike OpenAI, Microsoft, and Google, which have attempted to grow their B2C businesses through subscription models. The capital invested required to offer these services to customers is not going to be solved via subscriptions.
We do not speak enough about the applications that Amazon uses AI inside its logistics business to ensure that broken items are not shipped to customers, or to manage Amazon’s buildings. AI provides Amazon with the opportunity to further optimize its business and achieve profitability, but according to the press, Amazon appears to be struggling with AI.
I believe it is essential to note that Amazon is unlikely to acquire any B2C-focused businesses, as US regulators are likely to block such transactions. Investing in technology and offering resources it owns, such as cloud and computing power, essentially provides Amazon with an asset that it can sell at whatever price it wants and still make a profit on. I noticed during the last earnings call that some of their Antropic equity they converted.
I will reiterate a point I often mention – Amazon has been collecting data since its inception, and in terms of commerce, that should concern its rivals. Amazon will mitigate as much risk as possible regarding the collection of data to train their AI agents and tools on, but they will defend against startups and AI rivals such as OpenAI and others.
The Future
In conclusion, Andy Jassy and his management team will drive profits and uphold the vision set by Jeff Bezos. Amazon cannot lose in the AI race as it is the future. All businesses in the future will have less people working at them as more tasks are automated by technology. We are in the early stages of Day 1 for Ai. So when Jassy tells the Amazon workforce that the number of Amazon staff will be less in the future, he is stating what we are all seeing.
I would like to thank George Westcott for listening to this conversation for months and helping me fine-tune my thinking. I am still long on Amazon and interested in seeing what Andy Jassy and his team do in the short, medium, and long term.