Book Summary: The Innovator’s Dilemma (when new technologies cause great firms to fail) by Clayton Christensen

The book presents Clayton’s counter-intuitive thesis on how firms with good management practices and a sound understanding of their customers need, eventually fail at disruptive innovations while still succeeding at sustainable innovations.
The book emphasizes that its not engineering but management oversight that leads to the demise of incumbents (in the face of disruptive innovations).

One line summary: At some point, the incumbent’s product’s performance exceeds the demand of most customers and the “edge” which this performance metrics provided is lost, customers’ value proposition changes, they start valuing some other metrics, along which a disruptor’s product has better performance (and has early mover’s advantage), leading to demise of incumbent.
Following are the salient ideas raised in the book.

The Innovator's Dilemma

Sustaining Innovation: They improve the performance of established products, along with the dimensions of product that mainstream customers in major markets have historically valued. Established firms are usually good at sustained innovations.
Disruptive Innovation: They result in worst performance (along with those established metrics) at least, in near terms, they are usually cheaper (in net price, usually more expensive on per performance basis)m they start by capturing emerging markets.
Disruptive innovations are usually technologically straight forward. They are usually made by new entrants instead of incumbents.
Sustaining innovations are usually aimed at moving upward along certain performance metrics which the leading customers demand (higher margin products).

Impact of value networks

Competing theories

  1. Organizational hierarchy as an impediment to innovation: Since most big companies are organized into hierarchical subgroups, its difficult to make any change/innovation which can cause conflict among multiple groups, innovation inside the group has much lower friction. That’s why these companies succeed at sustained innovation and fail at disruptive innovation (since that does not fit well with organizational chart).
  2. Capabilities and radical technologies as an impediment to innovation: Big companies have accumulated tons of past knowledge (domain expertise) and when a disruptive innovation comes, it destroys the value of that past knowledge (author explicitly mentioned that this theory does not work in cases he studied).

Value Networks: People want to work on things, they believe are valued, most organizations value things which bring in more revenue, revenue comes from lead customers, therefore, lead customers determine the direction in which an organization invests its resources. Even when established firms innovated on disruptive technologies and their marketing personnel got negative response from lead customers, it forced the firms to focus on sustaining current innovation, this creates a room for new entrant to produce disruptive innovation and capture new/small customers and from there it moves upwards, eventually, this threatens established firm, who tries to fight back with similar product (which usually has lower profit margin than their previous product) but rather than capturing the new market (new entrant’s customers), it ends up being sold to the same leading customers leading to reduced profits and at best, defending a portion of prior business. Companies (or sales/marketing personnel) are tied to lead customers since they will generate most revenues (bonuses for salespeople) in the short term.

Market Risk vs Competitive Risk

There is always a market risk that a market for an emerging product might not develop but by not taking that risk, incumbents push themselves into competitive risk whereas late entrants, they face the risk of entrenched competition.

Commoditization (during sustaining innovation)

Once the product’s differentiated specification exceeds market demand, it loses its meaning an becomes a commodity since customers no longer value that specification (ashishb: One-time processor clock speed were seen as a benchmark but that’s not the case anymore, leading to the commoditization of processors destroying the value proposition of Intel).


  1. Lead customers effectively control the patterns of resource allocation in well-run companies => align (a portion of the organization working on disruptive technology) with the right customer
    1. Align whole of the organization to work on new disruptive technology
      Micropolis decided to move from 8″ to 5.25″ drive and its 8″ drive lost to competitors, sales had to refocus from mainframe customers to minicomputer manufacturers.  It was a painful move for the whole company.
    2. Align the same organization to work on sustaining as well as disruption
      Control Data Corp. (CDC) focused simultaneously on 14″ (sustained) and 8″ (disruptive), the engineers working on 8″ were regularly pulled to off to work on customer issues of 14″ drives and CDC failed badly at 8″ drive. To not repeat the same mistake, they assigned a new facility to the team for working on 5.25″  drive and captured 20% market.
    3. Create a separate unit (in a different facility) or spin-off or acquire a new business which will run as a wholly owned subsidiary – the best option
      Quantum (8″ and 5.25″ drive maker) created a spin-off Plus development Corp. with 80% ownership which focused on 3.5″, eventually, sale of 8″ and 5.25″ died down, and Quantum took 100% ownership of its spin-off to become the largest producer of 3.5″ drives.
  2. Small markets don’t solve the growth need for large companies => make a new unit (spin-off) small enough that small opportunity will be exciting for it
    Executives are focused on keeping the share price going up (since that makes employees stock options more valuable), if the prices go down, companies lose the most incentivized talent (which usually are future leaders).
    Therefore, there is a strong incentive to focus on increasing revenues to keep the investors happy. Directly going into small markets will not add to their bottom line, therefore, they have three options to deal with small markets.

    1. Try to make the small market grow at a faster pace – not possible, small markets simply cannot satisfy the near-term growth requirement of big organizations.
      Apple tried to do that with its PDA Newton in 1993.
      While 43, 000 Apple II in two years was seen as a success in 1979, selling 140, 000 Newton in two years (1993-94) was seen as a failure
    2. Wait for the market to become large enough (And face competitive risk)
      Seagate waited for 3.5″ market to become large enough and then was not able to get customers in Connor Peripherals (who was first getting sale orders from customers and then manufacturing those custom drives).
    3. Create a small unit focused on that small market or acquire one
      People want to work on projects which address the need of customers (and hence, impact organizations’ need for profit and growth), if the project lacks this characteristic then managers have to waste time explaining why the project should be provided resources. Best people avoid such projects and when things get tight, these projects are first to be canceled.
      Allen Bradley (AB) Company was leader in electromechanical (EM) switches (used in HVAC), as they saw (inferior in terms of ruggedness) programmable motor controls emerging they brought two companies (one full and one partial stake) and eventually ones, programmable control became standard, out of five EM manufacturers, only AB was able to maintain its lead.
      Johnson and Johnson follow a similar strategy of holding ~160 autonomous companies.
  3. The ultimate use of disruptive technology is not known in advance, first few attempts to find the right market is going to fail => iterative search for product-market fit is important
    The strategy for dealing with sustained innovation is execution while the strategy for dealing with disruptive innovation is learning and discovery (with more tolerance towards failure)
    Guessing the right strategy at the outset is, therefore, not important. What’s more important is to conserve resources for the second and third iteration.
    HP made 1.3″ drive and (with rigidity) targeted PDA market for it, PDA market flopped, game makers were interested in cheaper lower quality and lower capacity 1.3″ drive which HP was not interested in pursuing (or was too exhausted to put more resources towards it).
    Honda launched motorbikes in USA which were inferior to Harley and BMW (in terms of power and pickup), out of sheer frustration of its engineers (who started dirt biking to vent it out), Honda realized that their bikes were better for off-the-road conditions and people started ordering them for the same, it became de facto dirt bike.
    Regular retailers were not interested in them and eventually, sports retailers agreed to sell the bike.
    Honda also estimated to sell 0.5 million bikes annually while they actually sold ten times that (this shows they had no idea of the market as well as market size either).
    Harley tried to introduce similar bikes but its distributors were more interested in selling regular Harley bikes (more profit margin) and hence failed. (Disruptive technologies neither fit in models of established firms nor their distributors).
  4. Organizations consist of RVP (resources, values, and processes), latter two are pretty rigid => use resources of main organization but do not take its processes and values (after all, those processes and values are aligned for sustained innovation along the current path, they are thus, sub-optimal for disruptive innovation).
    Managers are great at figuring out whether an employee is a right fit for a project or not, they should put the same thought into the organization as well, an organization’s values and processes are good for a certain type of innovation, they can fail miserably at other.
    Only companies that succeed at disruption innovation is whose organization size matches the size of opportunity (which initially, is pretty small).
    Resources are usually easy to acquire and move across an organization.
    Processes are established so that, employees can do certain repetitive tasks in a predefined fashion (reduced friction and increases efficiency), since these processes are targeted towards sustaining innovations, they are actually a hindrance to disruptive innovation (which would require a new set of processes).
    Values are something which organization values, they are explicitly or implicitly communicated to all employees, something like “we only focus on products whose gross margins are more than 40%” or “we only produce products with quality about certain threshold” are hinderance since employees who have internalized these values cannot value disruptive products (which are usually inferior along these metrics). Big companies implicitly value only big markets, their size becomes a hindrance to small markets (which can later become big).
    Resources tend to be flexible and can be used for a variety of situations, processes, and values, by nature, are inflexible.
    A startup’s capabilities reside mostly in its resources, for its initial success to continue (As it becomes big), it has to develop processes and values, which will ensure continuous production of products.
    Therefore, for a startup addressing a new problem is about getting the resources right, for a big company, its capabilities (and disabilities) are more tied to its (rigid) processes and values.
    DEC (dominated mini-computers) had resources to succeed to micro-computers but its processes were more aligned towards a two-year cycle (where everything was made in-house), while most companies were outsourcing making of parts for micro-computers. Also, it valued higher profit margin business of mini-computer.
    IBM acquired Rolm (PBX maker) in 1994 but rather than let it run as a standalone unit, it decided to integrate (and hence, push its processes and values) onto Rolm and it failed badly.
    Cisco acquired Stratacom (ATM and WAN equipment maker) and let it run as an independent unit, it infused more resources for its faster growth.
    Microsoft needed to build Internet Explorer (IE) to compete with Netscape since Internet Explorer was a sustaining innovation (it strengthened its existing Windows product) but required a new process, a heavyweight internal team was needed (instead of the autonomous unit).
    Compaq in 1999, tried to sell computers over the Internet and its retailers forced it to back off (a different autonomous unit might have succeeded).
    Dell started selling computers over the telephone and later started selling them over the Internet was able to do that without any issues (since for Dell this was sustaining innovation).
  5. What makes disruptive technology unattractive for established market constitutes their greatest strength in emerging markets => find new markets which will value the product, ignore established markets as long as the product is not up to the established market’s demand.
    Product competition has four phases, the basis of competition changes from functionality, to the reliability, then convenience and finally price (commoditization).
    The real challenge for a new/disruptive product is a marketing one, it has to find a market where product competition occurs along dimensions that favored the disruptive attributes of the product.
    Thus, the aim should always be to provide adequate functionality (rather than superior since exceeding the market’s demand is redundant).
    Quicken (by Intuit) was a later entrant in the small business accounting software market but captured 70% market by producing a simpler product (eg. previous software required two entries per record, Quicken changed it to one).
    Genentech invested $ 1B to produce human equivalent Insulin (earlier ones were animal-based and a small fraction of humans built resistance against them) which were more purified, they tried to sell them at 25% premium and failed. At the same time, Novo (Danish Insulin maker) made Insulin pens which made injecting Insulin easier, it was able to sell its product for 30% premium (Genentech’s product was a commodity, while Novo was a premium product in eyes of customers).
    Sony built transistor-based radios which were of lower quality but customers valued there portability.

Some more examples

  1. Sears dominated departmental stores era but lost to discount stores, once the quality of products sold by discount stores improved beyond customer’s requirements.
  2. IBM dominated mainframe but failed to capture minicomputer, DEC dominated minicomputer but lost the microcomputer market. (Interestingly, IBM returned to dominated PC market for a while with its stand-alone PC division).
  3. As hard drives became smaller, at each stage, new entrants found new customers to sell to, the new markets (initially) were too small to be lucrative for incumbents, 8″ (customer – mini frame), 5.25″ (customer – mini), 3.5″ (customer – micro), 2.5″ (customer – portable computers), 1.8″ (customer – heart monitor and automobiles).
    Interestingly, 1.8″ drive was developed at a big company first which failed to market it to automobiles, who were buying them from a small startup instead.
  4. Mechanical excavators were more powerful than hydraulic ones (hence, more preferred for mining), so, the maker of hydraulic ones initially found a market in home construction (eg. for sewer lines). Eventually, “power” metric of hydraulic exceeded market’s demand (though remained less powerful that mechanical excavators) and people started using them in mining since they ranked better on metrics of “safety” (mechanical excavators were hazardous when a cable snapped).
  5. US integrated mill makers focused on the higher end of the market (like cans) while mini-mills (initially) produced lower quality steel and focused on (low margin) reinforcement bars, later, as they improved the quality, they moved to the higher end of the market.
  6. HP created a separate unit to focus on ink-jet printers, while the main organization focused on laser printers, they ended up targeted separate markets giving HP edge over Cannon.
  7. Intel was focused on DRAM market in the 1970s, its resource allocation formula allocated capacity proportional to gross margins of product lines, so, more resources were shifted towards microprocessor automatically and eventually without an explicit strategy, Intel became the market leader of microprocessor market (which Gordon Moore himself underestimated).

2 Replies to “Book Summary: The Innovator’s Dilemma (when new technologies cause great firms to fail) by Clayton Christensen”

  1. Great summary, thanks for writing.

    Another recap of Innovator’s Dilemma (in video format) that I recently found is here:

  2. Thanks for your effort and good mind.

Leave a Reply

Your email address will not be published. Required fields are marked *