Part 2 – Why Micro Four-Thirds is the Sensor of the Future: Switching Costs

June 3, 2013

Last week I explained why I believe a smaller format, probably micro four-thirds (m43), camera is in your future. Today I’ll tell you why that future camera probably won’t have the name Canon or Nikon on the body in 10 years.

Switching Costs

Canon and Nikon have used proprietary lens formats for years that impose high switching costs on consumers. Lenses built for one brand of camera will not work with full functionality on other brands; likewise, the camera bodies won’t work well with additional accessories like flashes. Once a consumer invests in a DSLR camera system, the cost to switch brands is significant and very few consumers will make the change. As long as the Canon and Nikon duopoly marches pretty much in lockstep, consumers will maintain brand loyalty even when one brand has a temporary strategic advantage in its lineup. It is cheaper to wait for the other brand to catch up, rather than switch for a fleeting improvement.

When I teach MBA corporate strategy classes I emphasize that switching costs are a significant barrier to entry for new firms in an industry. Established industry players often design products and services with these costs in mind to keep customers loyal (e.g. termination charges for cell phone contracts). Switching costs are a very effective strategic tool…until a disruptive technology emerges.

Disruptive Technology

Significant disruptive platform changes have happened infrequently in the photography industry. Large view cameras dominated the professional market until the emergence of the smaller platforms in the 1920s. Even so, professionals and serious amateurs used cameras like the 4×5 Graflex Speed Graphic for years.

New technology often spurs a flurry of innovation in the old technology. This was true with the Speed Graphic. Graflex introduced the Pacemaker Speed Graphic with multiple improvements to counter the emerging 35mm trend in 1947. Nevertheless, by the 1960s it was clear that the 35mm platform would become the industry standard. (See link to Gustavson’s book, Camera, at the end of this article for more on the history of the camera industry)

When new platforms emerge, this levels the playing field for new entrants. Investments in new systems require significant switching costs regardless of the sunk costs in the old system. This was the case for me with the move to m43. If I was going to move to a smaller format, my investment in Nikon glass was largely irrelevant. Moreover, the mediocre customer service and odd warranty policies of firms like Nikon and Canon mean that customers have little brand loyalty beyond the sunk costs of proprietary equipment.

The Canon and Nikon Reaction to Mirrorless

You might think that Canon and Nikon have all the technology necessary to dominate the smaller sensor mirrorless market. While this might be the case, it’s not likely to happen.

If established players reacted well to disruptive innovations, the world would look very different today. Wal-Mart would be leading electronic commerce instead of Amazon. Kodak would be leading digital imaging. Underwood and IBM would be leading word processing companies. Sony (of Walkman fame) would manufacture the leading mp3 player. Motorola would be leading the smartphone industry. The fact is, incumbent firms don’t tend to respond well to disruptive innovation. The economist Joseph Schumpeter’s description of the creative destruction caused by innovation means that incumbent firms tend to protect their existing turf rather than blow up their long-standing business model.

This willingness to embrace creative destruction has been one of the primary reason’s for Apple’s success over the past decade. Imagine if Steve Jobs had decided to exclude a music player for the iPhone to protect the iPod. Imagine if the iPad’s capabilities were limited by design in order to protect the MacBook. The willingness to risk a current line of products to an unproven new technology is rare in large corporate board rooms.

You can see the protective nature of the response of Canon and Nikon in their current mirrorless offerings. Canon is trying to preserve the value of its lens mount by introducing the EOS-M system. This APS-C offering uses existing lenses, but misses out on the size benefits of smaller systems like m4/3.

Nikon has gone the other way. Nikon has introduced its Nikon 1 Series that uses a smaller format than m43. This camera system has limited lenses and seemingly limited support in the Nikon corporate strategy. Moreover, the lack of a variety of fast lenses for the system makes achieving narrow depth of field difficult. The recent introduction of a 32.5mm f/1.2 lens for the system may mean this is changing, but it’s still early in the Nikon 1 lifecycle. The system does have advantages in autofocus technology over most other mirrorless offerings, so there may be some hope here. Nevertheless, Nikon’s mirrorless system seems to be positioned to not cannibalize from its larger format cousins rather than truly meet a market need.

The emergence of the m43 system and the entry of two “outsider” firms (Panasonic and Olympus) as early leaders fits the expected mold for disruptive technology. Panasonic is a relatively recent serious contender in the camera industry and has significant holdings outside the industry. Olympus is trying to reboot its camera offerings and is really a medical imaging company first and foremost. Recent alliances with Sony are an interesting development for Olympus. The universal lens mount for the m43 format has interesting implications for the proprietary business model designed to impose switching costs on consumers that has been prevalent for decades.

Of the two current industry leaders, Nikon has the most to lose. Nikon is not significantly diversified like Canon or the other industry firms. For Nikon’s long-term survival, it will need to become a leader in the mirrorless space.

If it seems far-fetched that I am predicting the demise of Canon and/or Nikon in the camera industry, consider this. Academic studies tell us that the average current life expectancy for large firms is 10 years and shrinking. You only have to think of a few examples to see how industry leaders can fall from grace quickly: Kodak, Circuit City, Netscape, America Online, Polaroid, Borders Books. Each story is different in details, but the overall trajectory is the same. Ten years from now, it’s very likely that either Canon or Nikon won’t exist as a camera manufacturer. You heard it here first.

Additional Reading

The history of the camera industry is captured beautifully in Todd Gustavson’s excellent book, Camera: A History of Photography from Daguerreotype to Digital. Every camera lover should own this book. You can purchase it at Amazon with this link:

Gustavson, T. (2009) Camera: A History of Photography from Daguerreotype to Digital

The classic text on the concept of creative destruction is at Amazon here:

Paperback: Schumpeter, J. (1950) Capitalism, Socialism, and Democracy

Kindle: Schumpeter, J. (1950) Capitalism, Socialism, and Democracy

Clayton Christensen wrote the definitive book on disruptive innovation. Amazon links:

Paperback: Christensen, C. (1997) The Innovator’s Dilemma

Kindle: Christensen, C. (1997) The Innovator’s Dilemma

6 thoughts on “Part 2 – Why Micro Four-Thirds is the Sensor of the Future: Switching Costs

  1. Judith Nicholls

    What you don’t take into account is the status factor of Nikon and Canon. To many (mostly males, I think) a large Canon or Nikon around their neck is like their BMW…

  2. Eamon Hickey

    Two quick thoughts on this:

    I think there’s an evidence selection issue with your argument — as there is with every argument I’ve ever seen based on the idea of disruptive innovation. There are also plenty of cases where established market leaders use their brand, manufacturing, distribution, technology and other advantages to successfully adapt to important new innovations. Two companies you discuss are perfect illustrations of this, quite recently and related to a much, much bigger market disruption than mirrorless ILC technology.

    In the transition from film to digital, which began in about 1996-97, Olympus and Sony jumped out to a huge early lead; the major camera company that was slowest to embrace digital was … drum roll, please … Canon. Fujio Mitarai (their legendary CEO of that time) later admitted that they had completely underestimated the impact of digital on the camera business. Yet, when they realized their mistake, they marshalled their quite formidable resources and in about two years (by late 2002) had taken over the lead in both compact and ILC digital cameras. Again, this was a disruptive transition that was far, far bigger than mirrorless ILCs. It killed Kodak, as you pointed out — but it didn’t kill Canon. One failure to adapt. One successful adaptation. Evidence selection.

    Another example from the camera business is the advent of the autofocus, autoexposure, autoloading compact camera of the late 1970s — a case study that, as I’m sure you’re aware, is famous in business schools. Konica shocked the camera industry with the first version of what we came to know as the AF compact. Turned the camera industry upside down. But within 5-6 years, the big players of the 1960s and 70s — Canon, Pentax, Nikon et. al. — were back at the top of the leader board, and Konica was on its way back to being a 2nd-tier player in the market (and now, of course, is out of the camera business entirely.)

    There might be a good book in this (probably already been written) — i.e. show the cases where the disruptive innovator failed to gain long-term market leadership; where the established leaders successfully adapted.

    Then, my second point is related to my first: I think you’re overestimating how big a disruptive innovation mirrorless ILC technology is. It’s not that big a change — just a variation on ILC cameras and using technology that every major camera company is already extremely familiar with (from building compact digicams). I think things like seamless network connectivity, programmability (a la smartphones), and seamless still/video convergence — if any of them, or all of them together, can be perfected — are all potentially much bigger disruptions. (The network connectivity/programmability combination is actually already a huge disruption in the camera business outside of interchangeable-lens models.)

    1. Michael Weeks Post author


      Thanks for your insightful comments on this post. I’ll try to give you some clarification to address the issues you raise.

      The first issue relates to your concern about selection bias. Selection bias implies that the evidence selected for examination is more likely to produce a particular result due to some internal characteristic or characteristics. I don’t think the m4/3 technology falls into this category. It very likely might fail. Its success is certainly not guaranteed and doesn’t seem to be anymore likely than other competing technologies.

      Nevertheless, I do acknowledge that there is a potential for bias that must be guarded against. Certain biases common in innovation research may be of more concern than selection bias. The first is a general pro-innovation bias that pervades much of the innovation literature. Pro-innovation bias refers to the concept that innovation is inherently good and beneficial for society or an industry. The problem for researchers with this bias is that often interview subjects (and researchers) don’t consider a technology innovative unless it is successful. Some research requires market adoption to be included in the sample. Innovations that fail to achieve market adoption are considered inventions. This is not a problem for some research questions, but can be an issue in some studies.

      A second and related bias is survivor bias. I think this is the bias that is really at the heart of your concern. Survival bias occurs when researchers examine a population ex post and miss part of the story because only the survivors remain to tell their story. This is a problem when exploring innovation topics. Failed innovation efforts often lead to market exit. This makes it more difficult to explore both sides of the equation. That’s one of the things I like about the m4/3 topic. We are not exploring it after the fact. We can watch it unfold. It has all of the characteristics of a disruptive technology. Will the Christensen model have predictive value?

      The second concern you raise is a good one. The big manufacturers have risen to the challenge for years, despite the fact that they are rarely first movers in the technology race. This is where you have to consider the finer points of Christensen’s research. While each of the technologies you mention (digital, autofocus, etc.) were disruptive in nature, the innovations moved from the most advanced models in the lineup to the lower cost models. Christensen’s framework typically identifies lower cost, lower performing innovations that appeal to a new customer base as the disruptive threat to established players in an industry. While disruptive in nature, the innovations you mention appealed to the established base of customers for the firms and were actually higher margin items soon after introduction.

      Fundamentally, I think your unease with the disruptive innovation concept is that most of the research on the issue has been descriptive and not prescriptive. We look at technologies after the fact and then identify them as disruptive.

      If the disruptive innovation concept is going to be truly powerful it must be prescriptive. The only way we will be able to advance our understanding of these issues (due to the biases mentioned earlier) is to identify technologies that fit the disruptive innovation model early in their lifecycle and follow their progress.

      As I said, the m4/3 has the characteristics that Christensen identifies in his work to consider it a disruptive technology (lower performance, lower cost, new customer base). Christensen’s model predicts that established firms faced with these innovation dilemmas often fail to react until it is too late. They cede the lower end of the market while protecting higher margin products until the market for the higher end products is too small to sustain the market and/or the firm. Will that be the case here? No one knows for sure. If Canon and Nikon respond and dominate the mirrorless, small sensor market, we can learn about the problems with Christensen’s model from following the diffusion of the technology. How were the firms able to overcome the tendency to protect established high margin markets and develop disruptive technologies? As you say, there’s a book in this.

      If Canon and Nikon fall into the classic trap Christensen identifies as the “innovator’s dilemma,” then we have ex ante evidence of the robustness of his model. That is, we were able to predict firm behaviors based on the market and firm characteristics he develops in his framework. This is actually less exciting from a researcher’s perspective, but given the descriptive nature of most of the previous work, it would still be a significant contribution to the field.

      Ultimately, the move to mirrorless is just an intermediate step along the way. I’m sure we’ll all be able to take “full frame” quality pictures with our cell phones in the future. In the mean time, we can enjoy watching the technology advance.

      Thanks again for your comments. I appreciate the conversation and hope you’ll continue to read and engage.

  3. Marco

    Very interesting and well done article. In general you could conclude that “open standards”, which require lower switching costs, will win. But in this case, I think m4/3 sensors, even as good as they are today (and even better tomorrow) won’t be able to convince everyone. I’m referring to full frame shooters and, to a lesser extent, to aps-c shooters. The difference in dr and depth of field is still quite evident, especially comparing ff sensors to m4/3 ones. But, possibly, there is space for something like m4/3 consortium with bigger sensors?

    1. Michael Weeks Post author


      It’s a difficult question you raise. The depth of field issue is one that technology is not likely to solve in the near term. I can see software solutions, but overcoming the laws of physics with the optical properties of lenses is unlikely.

      Nevertheless, I don’t find the depth of field to be a limiting factor for me with m4/3. It’s not as flexible as large sensor cameras, but I can work around it and get the results I like. I think the weight/size and dof tradeoffs are worth it, but I know that many will not agree. I was at a presentation by a successful fashion photographer last night. He said during his presentation that he rarely goes below f/4 because the focus becomes too difficult when the model moves as she poses. His work was beautiful and certainly no one would criticize him for not having a shallow depth of field.

      I’d love to see a consortium to offer “open standard” lenses for a new type of camera. The problem is that most of the market wouldn’t buy it. A new lens type for larger format sensors wouldn’t appeal to most people that already own Canon or Nikon gear. What would be the incentive to switch? The only way I can see a switch to an open standards approach is with a totally new platform, like m4/3, that has a differentiating factor in the market (in this case, size/weight).

      The dof field issue will remain for many traditional DSLR users. As I said in the first post of this series, if you’re after the ultimate IQ then m4/3 will never be sufficient. Micro Four Thirds will appeal to the new base of users that are moving up from cell phone and other compacts. It will also appeal to traditional users that find the IQ to be “good enough.” I think the smaller sensor, mirrorless market will continue to grow and large format DSLRs will become more of a niche product, much like medium and large format cameras are today. It will take time, but it seems likely to occur. The move to smaller formats has been going on since the 19th century; the pace just seems to be accelerating.

      Thanks for reading and thanks for taking time to post your thoughts.

  4. Louis Lin

    Except m43 (or any other sensor size) is not as strong as a platform as 35mm film.

    In the film days, there were a complex ecosystem associated with a particular film format. If I chose a particular film format, I had to make sure I could easily buy film, had it developed and had the right equipments to project the pictures. Film companies had to produce the right format of film in order to have it sold in volume. Same for camera makers and development labs. Everyone was interrelated through a common film size.

    This complex ecosystem encouraged a dominant film format (i.e. 35mm) to emerge. Once established, it was almost impossible for any single group (camera makers, film makers, consumers) to establish another dominant format (e.g. apsc film). All other film sizes become niche formats.

    Digital ecosystem is much simpler. Except camera bodies and lenses, different sensor sizes are compatible. I can use the same memory cards, same computer and the same labs to handle images produced by FF, APSC, m43 and other formats. For consumers. switching cost is just the cost to replace cameras and lenses. For camera makers, they don’t have to worry about consumers cannot handle a particular sensor size.

    No single digital sensor size can create a complex ecosystem like 35mm film. What is more likely to happen with digital is multiple sensor sizes will co-exist without a single dominant sensor size. m43 and APSC mirrorless will probably account for the volume, but there will be enough space for FF and smaller size ILC to have significant market share.

    So Canikon will not disappear in the foreseeable future. No doubt they will have less market share, but their current domination in FF DSLR will ensure their future dominance in the FF space.

Comments are closed.