Watch out! The new tech companies are eating your lunch!

Technology is the biggest sector in the S&P 500 index and accounts for five of the top ten largest companies. The sector is growing at a pace and scale never previously imagined and is eating into the profits of every other industry.

Tech giant Amazon is a prime example (no pun intended)!

But is it a retailer, a technology business or a logistics business? The answer is increasingly all three with a scale that is unmatchable.

In retail, Amazon’s scale is so staggering that it created out of nowhere the biggest shopping day – ‘Prime Day’ – when its prime members enjoy a one-day exclusive sale. On Prime Day earlier this month, Amazon sold over $0.5bn worth of merchandise, up 60% on Prime Day 2015.

In technology, its public cloud business (Amazon Web Services) is almost ten times larger than its next competitor, Microsoft, yet was launched just 10 years ago. With this business it essentially gave birth to the public cloud. Moreover, Amazon has introduced brand new product categories, such as its hands-free speaker Echo, ahead of Google or Apple and also offers a Netflix-like video service that has won several high profile TV awards.

In logistics, Amazon runs over 120 fulfilment centres around the world. I visited the centre in Hemel Hempstead in the UK and saw first-hand the use of data to optimise, create efficiencies and reduce errors on an incredible scale. The facility stores five million items and has around 1000 full-time employees, which means an employee manages roughly 5,000 items. Compare that to a typical supermarket employee who manages 500 items (source). Although not a like-for-like comparison, the magnitude of difference tells us that Amazon’s operation must be super-efficient.

These observations prompt the following questions: how can Amazon build such a large scale in such a short time and how did it become so successful in these seemingly disparate industries?

The answers to both questions are connected. First, it is the power of network effects that leads to the immense scale which makes Amazon’s support functions so large as to become standalone businesses. Second, the data-driven culture allows the company to run at scale efficiently and continue to make improvements.

Amazon started off as an online retailer of books. It launched an online platform which offered the world’s biggest selection of books in one place. That created a one-stop shop for anyone who wanted to buy books. It also created scale. Amazon utilised this at-scale online platform to extend into new categories of products – consumer electronics, consumer packaged goods and apparel. This platform grew so large that Amazon had to build enormous data centres to manage it.

Amazon then decided to sell the spare capacity to anyone who wanted to use it. This is how the public cloud business (Web Services) was born. As the retail business continued to grow, Amazon opened up its online platform as well as its warehouses for any third-party seller to use. Today, third parties sell more merchandise on the Amazon.com platform than Amazon itself. Opening up its logistics capabilities to third parties brings in a new, very profitable, revenue stream for Amazon – no inventory and potentially selling the logistics service at cost, but retaining a ‘platform fee’ for selling on Amazon.com.  And all of this while re-enforcing the value of its online platform to both its customers and more third-party sellers. That’s the power of network effects! (See following diagram.)

Amazon-tech-chart

Finally, to continue to attract new users and capture a higher share of their spending, Amazon has improved the value proposition of its platform. It offers a membership fee-based Prime service, which allows customers to receive their items within two days with no delivery charge. This leads users to order more often. I tested Amazon Prime usage with a straw poll amongst my colleagues on the equities floor (40 entries). I found that our average spend (Prime and non-Prime) is around £650 a year – a little more than Amazon’s average user spend – with around 30 orders each per year. On my sample, Prime users spent three times as much as non-Prime, because they placed three times more orders a year. Could it be that Prime is causing people to shift more of their household spending to Amazon as its range continues to broaden?

Underlying everything, I feel there is a very data-driven culture that is shared across all businesses, ensuring efficient operations. This means Amazon collects as much data as it can, uses that data to optimise as many decisions as possible, and continually improves the system by introducing new technologies.

So Amazon is a retailer, a technology business and a logistics business. It is a technology disruptor eating into various profit pools.

The company continues to add new categories – most recently groceries and food delivery – and in the longer term new technologies such as robots and autonomous vehicles/drones will improve distribution efficiencies. Eventually, could Amazon spin out these distribution services as standalone units? Most likely, yes, if history is anything to go by.

The world is changing and businesses across several industries need to accept that. Disruption by technology companies is the new reality, and Amazon isn’t the only one. Google and Facebook are altering how we spend our free time and, in doing so, capturing advertising dollars. Uber and AirBnB are disrupting the taxi/car rental and hotel businesses. Some of the ones just mentioned are re-inventing cars that drive themselves. Financial services are under threat from new FinTech start-ups that charge customers a fraction of what they have been paying.  Underlying all these disruptions are some major changes within three areas – consumers, technology and business models – that these technology companies play into, a topic I will delve into over the coming weeks.


The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested. Past performance is not a guide to future performance.