In 2018 Amazon became the second $1 trillion company. Less than 30 years after Jeff Bezos founded the online bookseller in Seattle, Amazon dominates e-commerce, is the world’s largest AI assistant provider and live-stream platform, and continues to be the top cloud computing platform, growing even more during the pandemic.

Clearly, if there is any company that can teach us a thing or two about strategy, Amazon is it! Not surprisingly, few MBA classes end without a student referencing Amazon. The same goes for business books, case studies and articles. Indeed, 1,451 Harvard Business Review articles mention Amazon, three times as many as Wal-Mart, which employs more people and generates higher revenues.

We love to learn from the best. Tempting but problematic! By being an outlier, Amazon almost by definition is able to do things few other companies can. Notably it breaks some of the fundamental rules of strategy. Amazon can do this, but for most others, the consequences will be dire.

Here are the three rules Amazon breaks, which you should probably stick to.

1 Strategy is about focus

One of the cornerstones of modern strategy is the idea that companies should focus on what they are really good at. “Stick to your knitting” is how Tom Peters and Robert H. Waterman Jr first put it in their 1982 bestseller In Search of Excellence

With growing deregulation, international competition and technology specialisation it is next to impossible to compete successfully in fundamentally different industries. That’s why Wal-Mart does not manufacture hair shampoo, Goldman Sachs does not sell insurance products and Jaguar Land Rover is not producing motorcycles.

Even well-run companies that try to venture into new industries typically fail. Allianz, for example, sold Dresdner Bank for half the price it paid seven years earlier with the intention of capturing a slice of the banking industry. Likewise AOL’s acquisition of Time Warner turned out to be a deal from hell, destroying $220 billion of shareholder value.

Amazon uses a version of the “razor and blades business model”. You won’t spend much on buying a Kindle but Amazon will make money from e-books you purchase on the device later on. What’s unique about Amazon is that the blades are from completely different industries.

At a 2016 technology conference Jeff Bezos put it the following way: “When we win a Golden Globe, it helps us sell more shoes.”

The point he makes is that Amazon thrives on customer interaction. The more people visit Amazon, the better. A Golden Globe nomination in other words brings them to the website but they end up buying something very different.

Amazon Prime is the best way to see why this works. If the only thing you get for $12.99 a month is free shipping, you are not likely to be tempted. So Amazon produces enticing TV series that you get with your Prime membership. Or it gives you access to discounted prescription medicine. Not only will you subscribe, you will also spend more time on Amazon, buying even more.

In their recent book Competing in the Age of AI: Strategy and Leadership When Algorithms and Networks Run the World Marco Iansiti and Karim Lakhani note that the new technology giants blur industry boundaries, thereby enabling massive scope increase. But this approach can not readily be adopted by companies that do not have a massive platform to leverage.

For mere mortals it makes sense to remember transaction cost economics. While moving into new industries will bring benefits in terms of new income, co-ordination costs (meaning the bureaucracy you need to manage the different businesses) will also rise. So rather than trying to manage very different businesses you are often better off buying products and services you need to make your own business thrive. Usually you are better off staying focused.

2 Don’t throw good money after bad

We are more affected by the loss of $100 than by the gain of $100. Behavioural economists Daniel Kahneman and Amos Tversky have shown that losses are twice as powerful as gains.

In corporate terms, this has profound implications as many executives find it difficult to cut cash-bleeding activities. It would, after all, force them to admit they have made the wrong decision. So instead, good money is thrown after bad in the hope of turning a business around.

At first glance Amazon seem to be subject to such loss aversion as well. In 2007 it launched a grocery delivery business that is expanding city by city.

Groceries are low margin but that has not stopped Amazon. In 2017 it paid a 27 per cent premium for the acquisition of Whole Foods, a brick-and-mortar business. The acquisition did little to boost Amazon’s grocery business. The company remains a small player in the US market. Will that prompt Amazon to finally give up? Not at all. We purchase groceries twice a week while getting a new laptop, for example, is a pretty rare event. So keeping groceries in its portfolio is primarily a move to get customers to its website.

As long as other companies understand that some activities are loss-leaders necessary to keep the core business booming, that’s fine. The danger is, however, that many justify loss-making activities by pointing to Amazon, thinking that it makes such investments primarily in the hope of turning its grocery arm into a highly profitable business. That’s loss aversion, the Achilles’ heel of many companies.

3 Your core competences determine what you can and can’t do

When Amazon started building the Kindle the company had no experience in how to develop and manufacture hardware.

“Originally I told Jeff (Bezos) it would take us about 18 months to build the Kindle and we could do it with a couple of handfuls of folks,” recalled Steve Kessel, who lead the original Kindle team. But “it took us three-and-a-half years and a lot more than a couple of handfuls of folks,” he conceded.

This is not untypical for how Amazon works. The company works backwards from what it sees as enticing offers for its customers and builds the skills needed.

Many companies try similar approaches, identifying opportunities, not worrying about their existing resource portfolio too much. But few companies have the deep pockets needed to build an entirely new set of competences.

Even those who have the cash rarely succeed in building new radically different resources, as Google’s ill-fated attempt to build a phone showed. In contrast Waymo, its successful play in self-driving cars, concentrates on software and leverages its competences in AI and big data. So instead of building hardware (where Google lacks competences) it developed deep partnerships with the likes of Fiat Chrysler.

Amazon’s notion to ‘work backwards and scale forward’ is not easily adopted by other companies. It’s better to engage in opportunities where the firm can leverage its competences.

And then there is luck

Amazon is a fascinating company but drawing lessons from its strategy is not straight forward. What’s possible for Amazon is not necessarily possible for any other company.

It actually poses the question whether we can learn from outliers at all. Chengwei Liu, a former Associate Professor at Warwick Business School and now at ESMT Berlin and author of Luck, is skeptical.

“Having studied Nascar drivers, hit movies, patent citations, and both growth and profitability of US companies, I found that learning from the top performers is problematic," said Dr Liu. “Luck rather than skill seems to determine who comes out on top.”

In other words, Amazon might have just been lucky. Liu's research certainly reminds us to stop and pause prior to applying lessons from outliers. They can be helpful to highlight extreme ideas, but how you can use them in your own business requires careful consideration of the differences.

This article is republished from Forbes.

Christian Stadler is Professor of Strategic Management and teaches Strategic Advantage on the Distance Learning MBA and Executive MBA.

Follow Christian Stadler on Twitter @EnduringSuccess.

For more articles on Strategy sign up to Core Insights here.