Define and refine your Core Strengths
This seems obvious. It’s so obvious that it is a strategic error that many companies make. First, they over define what are their core strengths. Defining a core strength is simple: any core strength must differentiate your product. For something to be a core strength, it must add value in a way that competitors can’t or are not good at doing.
A good test of your core strengths is to ask: do you view it as a cost center? If the answer is yes, then it cannot be a core strength. As second test is to ask: Would you be better off outsourcing it? If the answer is yes, then it cannot be a core strength. If the answer to both is no, then ask: Will you continue to invest in refining and strengthening your core strengths? If the answer is no and the answer to the previous two is yes, then it definitely cannot be a core strength. If the answer is yes and the answer to the previous two is no, then you will lose this core strength as your competitors eventually learn how to copy it at lower cost.
The most common source of this becoming a strategic error in tech is when either accountants or technologists dominate a company’s leadership. This is because accountants lean towards viewing everything as a cost center while technologists lean towards viewing everything as a core strength. Striking a fine balance between the two is one of the reasons why tech companies like Intel, Lam Research, and Applied Materials have done well over decades.
Many believe that TSMC’s Foundry model won over IDMs because it was fundamentally superior. Proof that this is a conceptual error lies in the fact that at least one Logic IDM and plenty of Analog and Memory IDMs continue to thrive. After almost three decades of Foundries, the latter should be long gone.
The reason why some IDMs continued to thrive is that they continued to refine manufacturing as a core strength. There are a group of IDMs that once dominated the semiconductor landscape in 60s, 70s, and 80s who became also-rans or disappeared from the competitive landscape. At the center of their systematic decline is the fact that they began to threat manufacturing as a cost center. But they failed to act on this belief, blurring the lines between their core strengths and their cost centers. This strategic error created a tactical error in that they were trying to compete against fundamentally lower-cost Asian Foundries from high-cost manufacturing bases. This financial disadvantage meant that they were dying the death of a thousand cuts.
In contrast, TSMC strategically treated manufacturing as a core strength, while it treated the design and marketing of finished devices as a key weakness. Clarity with these two led it to start with a strict rule that it would not compete with its customers. Understanding that design and marketing finished devices from Taiwan could only be done at a cost disadvantage meant that it could more openly partner with Fabless start-ups who could not afford a fab. In so doing they overlaid each other’s core strengths on the other partner’s cost centers,* thus creating a far more power combination.
The Foundry-Fabless partnership proved far more powerful against IDMs with far greater resources, only when those IDMs had blurred the lines between core strengths and cost centers.
So what about when those weakened IDMs went Fablite and continued to fail? The answer here is that Fablite was a continuation of their blurred lines between core strengths and cost centers. The Fablite model stopped investing in what they still believed was a core strength, as they continued to believe their existing manufacturing and development capability was a core strength. For many, these fabs would eventually be shut down as they lost their ability to produce value. Texas Instruments was one that did not. The reason is that they took their leading-edge digital capacity and shifted it to differentiate their Analog-Mixed-Signal devices. In the end, it starts with a deep assessment of your core strengths and cost centers.
* Cost centers are also often core weaknesses.
