In this blog we will touch on a few aspects of how to attack an entrenched competitor, a topic of interest whether you are playing offense or defense. Winning business attack strategies are difficult in this case. There are whole books on the topic so we will hit only a few points. Let’s start with a quote that captures the situation:
“The first and most important rule to observe. . . is to use our entire forces with the utmost energy. The second rule is to concentrate our power as much as possible against that section where the chief blows are to be delivered and to incur disadvantages elsewhere, so that our chances of success may increase at the decisive point. The third rule is never to waste time. Unless important advantages are to be gained from hesitation, it is necessary to set to work at once. By this speed a hundred enemy measures are nipped in the bud, and public opinion is won most rapidly. Finally, the fourth rule is to follow up our successes with the utmost energy. Only pursuit of the beaten enemy gives the fruits of victory.” – Karl von Clausewitz (see On War), (Aside: the fourth rule is aligned with our analysis of learning curves.)
In business, the attacker always gets to pick where they want to engage. Perhaps the attacker will fail in execution or make the mistake of playing the defender’s game instead of theirs, but let’s talk about how it is done when it is done right, which is to attack where the entrenched competitor is weakest and where the attacker can mass a superior force, as taught by von Clausewitz’s second rule. It does not matter if someone has 90% world wide market share if you can find a niche where they are structurally weak. As attacker, you pick the time and the place. The place you should pick is somewhere where the competitor cannot easily defend and where you can therefore have the larger force at the point of engagement. The defender has the problem of defending their whole franchise, supporting their installed base, worrying about finding solutions with sufficient generality to avoid product model proliferation, explaining their gross margin to Wall Street, etc. As attacker you are focused on growth.
It is important to pick the point with care because the defender has a lot of resources that they can, and will unless they are asleep, bring to bear. For instance, a common technique for the defense is to try to freeze the market. “Oh, we will have that, better integrated and with more features, in six months. Would you like to be a beta?” Customers almost always wait six months for the incumbent. So on offense you need to find a place (meaning a market segment) that the competitor cannot easily defend. The most classic of these is a low price point where, to defend against you, the competitor must discount their whole product line. But there are lots of problems with this unless you have an intrinsically lower cost structure, for instance, it is hard to earn enough to support your R&D and, since it is unlikely that you will have all the features that the entrenched competitor has, you have to go after a niche. (With each generalization, there are counter-examples. For instance, when introducing the iPod, Apple focused on carefully increasing the feature set so as to have a whole product solution.)
If it is a niche then it can, by definition, support a different price point and the defender can often just create a model variant of their product that they can compete with you on without hurting the rest of their product line. We talked about how to do that earlier. Of course, the defender could choose to abandon the market because it is unimportant and defending would hurt their average gross margin, but most companies know better than to do that—they will do their best to deny you the pure oxygen of uncontested revenue. A better technique is to have some technical or IP (intellectual property) differentiation that matters to the customer. In high-tech, it does not really matter whether you lead with a technical idea or with a customer problem, in the end you need both technical (or process) differentiation (or it will be too easy to copy) and customer intimacy (or you will miss the mark and give the defender time to respond). [Many books say that marketing is more important, but I think that is because marketing guys are better at marketing.] Remember, differentiation equals gross margin.
I think about two large classes of attacks. The first is the extremely rare Clayton Christensen type of discontinuity. This is where a product used in a less demanding market segment attacks an entrenched incumbent from below. This is a devastating attack because it is so hard to defend against. On the other hand, it is very rare. But the book is so popular, and it is a great book, that this attack form has taken what I believe is a disproportionate amount of strategic attention. Read the books [1
When you attack with a highly differentiated product, the first thing to realize is that adoption could take longer than you think. It could take a decade. In the industrial market, the industry is likely to go through at least two downturns before your business is a healthy ongoing concern. One down turn is a virtual certainty. So you need to size the business (spending and upside) so that it doesn’t get killed off in the first downturn. For instance, the vast majority of product lines in the semiconductor capital equipment industry don’t make any money. They are sometimes profitable in the upturn but lose that money in the downturn. This is why getting the enterprise going with an acquisition can be a desirable strategy for large companies. Unfortunately everyone knows this so many businesses are overpriced by M&A speculators, making an acquisition also a risky bet in terms of shareholder return. So be careful when you size the new business and commit to a fixed cost structure. You will also want to make as much variable as possible. You will be able to drive out some of these inefficient costs later, should you live so long. Since you will have to have a slightly more expensive cost structure so that you can vary it more, you will find this yet another reason not to compete only on price unless you have a fundamentally lower cost structure (think China and India).
What everyone rightfully looks for in a place to attack is a discontinuity. Discontinuities happen all the time in the semiconductor industry because we move so fast. Examples include aluminum to copper wiring, bipolar to nMOS to CMOS logic, CMP, 365 nm to 248 nm lithography, chrome on glass to embedded phase shift masks, 200 mm to 300 mm wafers. Today, leakage currents are tremendously important whereas a few years ago they were of negligible importance for mainstream applications. The challenge with discontinuities is that they take a long time. Every technology can be extended through incremental improvement long past when most people predict that they would expire. The most famous example of this is optical lithography, whose demise was hinted at in Gordon Moore’s first paper on Moore’s Law four decades ago. I once read an interview with the founder of a company that was an “overnight success.” It had taken him 20 years to achieve this status. Nevertheless, discontinuities or “strategic inflection points” are the place to be. Many companies don’t even try to be first, but rather wait to see the market start to develop and then be a so called “fast follower.” We have a perception that being first is best and that the first entrant typically captures the market and keeps it. Intriguingly, detailed studies of the facts show that much of this early mover advantage is simply the victor re-writing history so that they were “first” when in reality they were not. Of course as the market matures and the discontinuity is history, it becomes quite difficult to displace the leader. Discontinuities happen with lower frequency in non-semiconductor industries, but they occur almost everywhere. The internet has caused discontinuities in many industries.
Looking at it from the customer side, one needs to find an unmet need. That need might be obvious to the customer or latent, but unless there is an unmeet need, why would anyone buy? People don’t like to change. Let me say that again—as a generalization, only a very few early adopters will want change. (The danger is that these folks will be those interacting with you so you will get a distorted and overly optimistic view of how fast the masses will adopt your solution—read Crossing the Chasm by Geoffrey Moore. Pay particular attention to the issues surrounding the “whole product.”) In order to capture most of the market people are going to need a compelling reason to take a risk on the new product. The incumbent that has been in the market a while has learned to segment the market in various ways and has optimized his business around that structure. That optimization is his strength and his weakness. You will probably want to find a market segmentation on which that the incumbent has not focused. That is a great way to start to find unmet needs.
Marketing Sirens: Of course if there is a large entrenched competitor, your customers will be very vocal in encouraging you to develop a competing product. Customers have many “needs” and a big one is to lower the price of the product of the market leader. You have to do enough market research to uncover whether it is the functionality or the price of the product of market leader that is the “customer problem.” In the semiconductor equipment business, it is also very common to be earnestly entreated to develop products that are needed for the customer’s backup plan. Sometimes the slow adoption of an industrial product is simply because the customer only wants that product to exist as a backup. In a reticle inspection business I once ran, if we had promptly invested heavily in every one of these our customer asked for—230 mm masks, alternating Phase Shift Mask database inspection, 157 nm hard pellicle inspection, EUV inspection, etc.—we would have been a lot less profitable. Intel has a formal two-wavelength strategy for lithography, which means that the area is so critical to them that they bet on two possible futures at any given point of time. Let me restate that; it is their corporate policy that 50% of what they tell you isn’t true, even though neither you nor they may know which half! If you pick the wrong one, you lose. If you bet heavily on both, given limited resources, you also lose. This is not Intel’s problem, they are behaving rationally, even brilliantly. Let the seller beware!
Since a wakeful market leader will respond to an insurgent the minute they get credible intelligence as to their plans, secrecy in the attack plan is paramount. Since you will need to work with customers to develop a (whole) new product, you need to find a way to protect the information both inside the company and at the customer. Compartmentalize information to the few people who have a need to know. With the industrial customer, company-to-company non-disclosure agreements (NDA’s) are generally not enough in my experience (I hope yours is better). I prefer special personal non-disclosure agreements that each person involved at the customer must sign. Nothing is perfect, but this is pretty good. Obviously, you have to choose a group in the customer base that wants you to win, but you need to keep in mind that it is in the best interest of the customer to socialize your secrets in order to get the best features at competitive prices from all of their vendors. It may not be ethical or legal, but those are the natural forces, so relationships are critically important. The customer needs to have in place a way to compartmentalize the information in their organization. Disinformation can also sometimes be used, but be careful. In the choice of partners to work with, look for people that can make you better, that will influence other buyers, that know what they are doing, that do not have an internal group in competition with you, that can keep your secrets, and that tend to go down the mainstream. You already know the people who are the worst at keeping your secrets—they are your best sources of information on your competitors. Still, don’t get too hung up on secrecy—the incumbent will often react slowly.
The size of the market segment you pick has a lot to do with your chance of success. There will be a certain amount of engineering needed to develop a steady stream of products. How much of the market do you have to get to become self-supporting during the upturn? During the downturn? Beware of picking segments that are too small. In the semiconductor equipment market, be cautious if the system is potentially independent of the technology node, as it is easy to saturate the market. You ideally want a product that becomes obsolete as linewidths shrink. There are lots of market segments that are intrinsically difficult to obsolete, such as the famous “pet rock” product of the 1970s. What do you do next?
In the semiconductor business there are buying cycles for nodes. If you miss the node, you are out. The other side of this is if you pick a technology for insertion at a node and it gets pushed out to the next node, which is happening more and more frequently, you have to wait another two or three years, letting the incumbent have more time catch up while you burn cash and try to defend your program. I have a sense that predicting technology insertion is getting more risky, not less, because of more emphasis on materials and 3-D device structures.
A good discipline is to put together the “elevator speech.” The scenario is as follows: you get on the elevator in the basement and your biggest potential customer gets on at the first floor. She pushes the button for the 22nd floor. (You pushed 14 but you are not getting off there, right?) You have 21 floors to clearly explain your compelling value proposition. If it is valid, but 55-floors-complex to convey, you have a big problem.
Let’s close with two military quotes:
“the art of concentrating strength at one point, forcing a breakthrough, rolling up and securing the flanks on either side, and then penetrating like lightning deep into his rear, before the enemy has time to react.” – Field Marshal Erwin Rommel (see Knight’s Cross : A Life of Field Marshal Erwin Rommel)
“the most consistently successful commanders, when faced by an enemy in a position that was strong naturally or materially, have hardly ever tackled it in a direct way. And when, under pressure of circumstances, they have risked a direct attack, the result has commonly been to blot their record with a failure.” – Sir Basil H. Liddel-Hart (see Strategy, History of the Second World War
And finally, from the oracle of Omaha, “When a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact.” – Warren Buffett (see The Essays of Warren Buffett : Lessons for Corporate America, The Warren Buffett Way)