Verticalization as an enabler of customer experience
Christmas did come early for us last year in the form of a brand new M1 Mac Mini that has replaced an ageing iMac. I made my concerns around Apple’s ecosystem vocal (here and here) and I still stand by that. However, I think Apple is (almost) single handedly changing the computing landscape as we know it with these new chips — or rather, it is the first personal computer manufacture to do so (I am not counting Microsoft’s half-hearted attempts — Apple is all-in).
There has been a quiet revolution happening with Arm chips and more recently with RISC-V. In my last blog post about the M1 Mac’s, I suggested that all the big tech companies are creating their own chips — but Apple is the only one with significant consumer space market share which for us mere mortals is an opportunity to see behind the curtains or in other words we can measure the difference between the Intel chips and Apple’s own.
I had written a whole paragraph around chip development, particularly Arm and RISC-V computer chips but then decided that it was too technical and not actually relevant for the bigger picture. Instead, I would like to shed some light on what I believe is happening here more broadly with changes in the value chain and in fact, I perceive the ability to design one’s chip and write the software for it as a direct consequence of the broader verticalization trend. The key takeaway is that any company that utilizes computer chips in its product line-up is confronted with the following constraints: power consumption, heat generation and cost. These three factors will generally determine how well the designed function, for which the computer chip is needed, can perform. And here we have the first lynch pin: most computer chips are general-purpose or in other words, they tend to be designed to be jack of many trades because chip companies such as Intel and AMD need to sell a certain volume to recoup the ROI. Of course, there are certain chips that focus on low-power consumption (these typically go into laptops and other type of appliances), but overall chip design and manufacturing is so expensive that demand fares better with a general-purpose chip. To make up for possible gaps that these chips have, additional components are added to the product (making it more power-hungry and less efficient in the process). A good example of this is the laptop or desktop you are using right now (if you are indeed). In short, there is always a compromise if you must deal with general-purpose computer chips and that comes at the expense of the customer experience.
Apple was not going to take it anymore, in fact Johny Srouji noted in an interview: “We want to create the best products we can” and added “we really needed our own custom silicon to deliver truly the best Macs we can deliver.” and if you read the sources, Apple wanted a different chip and no longer the compromise. Intel simply could not or would not deliver. So, they designed their own chip (not just for the new Mac’s they have done it since the A4 in the iPhone). Apple is not alone here, for example, Samsung and Huawei design their own computer chips for their smartphones; likewise, as mentioned in my previous blog; Tesla has created their own computer chips for their artificial intelligence application (they moved away from Nvidia); similarly, Amazon and Google have their own chip designs and finally, it is rumored that Microsoft does the same.
Underlying this development is the need for not just the best computer chip but one that is designed for the task at hand. I have read some articles that already talk about the (Innovator’s dilemma type-) disruption of the computer chip industry, but I think this is different because the disruption is the enabler not the cause. While you could argue from Intel’s point of view, it is clearly a threat when the above-mentioned companies create their own chips, but they are not in the business of computer chips and they do not compete with Intel. They simply create the best tool for the job, and it is all under the umbrella of forced backwards vertical integration. I say forced because they cannot get what they want from their supplier.
Let us come back to Tesla — Tesla not only designs everything, but they also produce most of their products in-house. When it comes to computer chips, Tesla has found itself in a similar situation as Apple. They simply could not get the exact feature set they wanted from Nvidia and so they designed their own chip for self-driving. By doing so, they control every aspect of the self-driving experience. But Tesla did not stop there because they encountered the same issue with battery technology and so they designed their own battery which they also intend to produce. This is summarized by Elon Musk as follows: “We’re designing and building so much more of the car than other OEMs who will largely go to the traditional supply base and, like I call it, catalog engineering. So, it’s not very adventurous and it basically ends up like older products end up — looking the same because they’re going to the same suppliers.”
But when we look at who started this trend, we clearly find Amazon at the helm. From its humble beginnings, Amazon has vertically integrated anything and everything. Today, you can order an Amazon tablet from Amazon which is then packaged and shipped by Amazon fulfillment and delivered (in some countries) by Amazon. When you receive the tablet, you can view an Amazon produced movie using Amazon’s own video player that streams from AWS (Amazon’s cloud infrastructure). Of course, Amazon has been dabbling in so many different areas that it requires another whole blog post to illustrate this.
But why is this happening now? In one word: customer experience. This is a broad term for which Harvard Business Review has a well written article on the philosophy and methodology; What it ultimately translates in the above examples is to create the best product or service for your customer, you must ensure control of all experiences. To be clear, I am not saying that customer experience is what drives verticalization, I am saying that these companies are forced to vertically integrated because their design of the customer experience cannot be achieved in any other way.
Apple has gone to the extreme with this philosophy and so did Tesla. Jeff Bezos has been clear about his all along: “If there’s one thing Amazon.com is about it’s obsessive attention to the customer experience, end-to-end.” Of course, these companies have something else in common, their all-big tech with deep pockets that can afford to do this.
I am more interested in understanding the mechanism in making this level of verticalization work for everyone. And truly, it should not require that you must own a chip- or battery design unit in your company. What is required is a certain level of supplier configuration and the pursuit of the best customer experience. Let us go back to the example of Intel (and probably AMD as well) where they need a certain volume of chips to cover the ROI. James Allworth wrote a great article to sums up Intel’s misstep; Yes, it’s referencing Clayton Christensen and the article talks about disruption and in the case of Intel it was in fact the decoupling of chip design and manufacturing, but really I want to zero in on this citation that is referenced from stratechery.com: “the company [Intel] refused to focus on power efficiency and preferred to dictate designs to customers like Apple, contemplating their new iPhone, instead of trying to accommodate them (like TSMC)”. In other words, if your supplier dictates the design (and with that the constraints to the experience), you will not be able to achieve the desired customer experience.
I am not advocating that everyone should now abandon their suppliers and vertically integrate; what I am advocating is that a) companies need to understand the significance of the customer experience, b) why big tech has jumped on this bandwagon in such extremes and c) how important the design of the customer experience is for customer satisfaction.
In short, assess the resulting experiences on customer engagement and determine if you have sufficient control on the experience. Then the question of verticalization or changing the supplier will be much easier.