Managing Innovation: Trends - Innovate or Buy

Essential Definitions (again) The “functional” source of innovation depends upon the functional relationship between innovator and innovation: – An innovation is a USER innovation when the developer expects to benefit by USING it; – An innovation is a MANUFACTURER innovation when the developer expects to benefit by SELLING it. We will cover two issues today: Two major factors that affect users’ decisions to innovate or buy z z Transaction costs The number of users that want the same product The kind of innovations that users develop in their “low cost innovation niches” is affected by: z Sticky information z Learning by doing Transaction costs z If you have the money to pay for a just-right product made just for you – can you find a manufacturer willing to build it for you? – If you want a “just-right” auto, should you call General Motors? What will they say if you do? – If you want a “just-right” auto, should you call a custom car builder? Will they be willing to make exactly what you want? What are agency costs? Agency costs are: (1) Costs incurred to monitor the agent to ensure that it follows the interests of the principal, (2) Cost incurred by the agent to commit itself not to act against the principal’s interest (the “bonding cost”), (3) Costs associated with an outcome that does not fully serve the interests of the principal. Agency costs induce manufacturers to not quite build what is precisely right for you z Users want to get their precise need specification met – because that is a lower-cost solution for them. – For example, a tennis player wants to get a racket that works better with his or her already-ingrained hitting technique. A precisely-right racket is cheap relative to relearning how to hit. z Manufacturers want to apply a solution that they “have in stock” and understand well – because that is low cost for them. – A metal tennis racket maker will not want to make you a carbon-fiber racket – even if carbon fiber would be a better solution for you as a custom buyer – but they won’t tell you that! – A disk drive maker will not want to make you a semiconductor memory, even if that solution fits your needs better… Another agency cost. Manufacturers must invest in quality signals where users don’t – Helping users to sometimes make cheaper solutions for themselves z A user can put a cheap part into an expensive product they own and know the quality is good. “Look, I saved $200! I replaced my BMW’s leaking window washer fluid tank in 5 minutes with a recycled plastic milk bottle. It works perfectly! z A manufacturer can’t. “BMW factory special! – 7 Series window washer fluid tanks are now made from used plastic milk bottles! – Engine electronics are now from Lego Mindstorms kits instead of Bosch! – Factory tests show both work perfectly! Model shows users can always make a unique product more cheaply for themselves than a custom manufacturer can make it for them when there are transaction costs – and there always are! z Assume both user and manufacturer are hiring from the same pool of designers to design the custom product. – Vij is the value of a solution to problem j for user i. – Nj is the number of users having problem j. – Whj be the cost of solving problem j, where W = hourly wage and hj = hours required to solve it. – Pj is the price charged by a manufacturer for a solution to problem j. – Let T be fixed or “setup” transaction costs, such as writing a general contract for buyers of a solution to problem j. – Let t be variable or “frictional” transaction costs, such as tailoring the general contract to a specific customer. Portion of Baldwin – von Hippel model of innovate or buy (See Chapter 4 in Democratizing Innovation) z A user has no transaction costs in dealing with itself, so a user’s payoff for solving problem j will be Vij – Whj. Therefore, a user will buy a solution from an upstream manufacturer rather than develop one for itself if and only if Pj ≤ Whj. z Assume t = 0 but T > 0. Then, the manufacturer’s payoff for solving problem j will be Vij – Whj, which needs to be positive in order for the manufacturer to find developing the innovation attractive. z Nj Pj – Whj – T > 0. But, as we saw, Pj ≤ Whj if the user is to buy, so we may substitute Whj for Pj in our inequality. Thus we obtain the following inequality as a condition for the user to buy: Nj(Whj) – Whj – T > 0, or Nj > (T / Whj) + 1. z In other words, Baldwin and I find that the absolute lower bound on N is greater than 1. This means that a single user will always prefer to solve a unique problem j for itself (except in Coase’s world, where T = 0, and the user will be indifferent). If every problem is unique to a single user, users will never choose to call on upstream manufacturers for solutions. It pays users to innovate rather than buy at the leading edge of markets – where N is small! When N > 1 due to high transaction costs, user communities are the cheapest innovation solution Individual lead users innovate here User COMMUNITIES can innovate more cheaply here Manufacturers can innovate more cheaply here # of users perceiving need Time Data supports the model: Users aren’t always the innovators Innovations Samples: User Manufr Suplr Other NA Total (N) Scientific Instruments 77% 23% - - 17 111 Semicon & PC Crd Process 67% 21% - 12% 6 49 Pultrusion Process 90% 10% - - - 10 Tractor Shovel Related 6% 94% - - - 11 Engineering Plastics 10% 90% - - - 5 Plastics Additives 8% 92% - - 4 16 Industrial Gas-Using 42% 17% 33% 8% - 12 Thermoplastic-Using 43% 14% 36% 7% - 14 Wire Stripping Equip 25% 75% - - 2 8 Connector Attaching Equip 4% 13% 83% - - 12 Sports Equipment 58% 27% - 15% - 48 User and Manufacturer Innovations Differ in ways that information asymmetries would predict Users tend to develop Functionally Novel innovations: z The first sports-nutrition bar z The first scientific instrument of a new type Manufacturers tend to develop Dimension of Merit Improvements: z A better-tasting sports-nutrition bar z Improvements to an existing type of scientific instrument Sticky information leads to information asymmetries and additionally affects who develops what Information is often very “sticky.” Some reasons: z Information needed by developers may be tacit – Can you tell your child how to ride a bike? z A lot of information is often needed by developers – “You didn’t tell me you were going to use the product that way!” To develop a product or service, information about needs and about solutions must be brought together at a single site. z z Need information is usually found at user sites. Solution information is usually found at manufacturer sites. Software Supplier Software User Solution Information Need Information Studies show this effect clearly Sample of 24 inventory control system innovations by Seven-Eleven Japan and NEC Stickiness of Technology Information Stickiness of User Need Information - + Amount of Technology Design Done by the User Amount of User Need Design Done by the User + = positive influence; - = negative influence Product or service design tends to move to the site of the crucial sticky information Manufacturer-Based Design (DOM products) Manufacturer design tasks • Have solution information • Acquire need info from user • Design product User design task Need Info Source User-Based Design (Functionally novel products) Manufacturer design task Solution Info Source User design tasks •Have need information • Acquire solution information •Design product Example of the impact of sticky information on the locus of innovation: Fifty percent of all prescriptions written in the U.S. are written for “off-label” uses of prescription drugs – New prescription drugs are generally developed in the labs of pharmaceutical firms – sites where much specialized information about drug development has been build up over the years. – Off-label applications are generally found by patients and physicians. They apply the drugs many times under widely varying field conditions – and discover unanticipated positive (or negative) effects thereby. (“Doctor: this blood pressure medication you gave me is causing my hair to regrow!”) Because information distribution is asymmetric, user innovation is widely distributed – a few “super users” do NOT do all the innovation Few users developed more than one major commercialized innovation Number of users developing this Number of major innovations Innovation Type 1 2 3 6 na (n) Scientific Instruments* 28 0 1 0 1 32 Scientific Instruments** 20 1 0 1 0 28 Process equipment*** 19 1 0 0 8 29 Sports equipment**** 7 0 0 0 0 7 Linus’ Law z More users find more bugs - because adding more users adds more ways of stressing the program. z More users make bugs easier to solve – “shallower”- by increasing the probability that “someone’s toolkit will be matched to the problem in such a way that the bug is shallow to that person” Learning by doing adds to local information economies to create narrow low-cost innovation niches Users have the advantage of problem-solving in their own use environments as they “do” a desired activity – they are “learning by doing.” Examples: – Airlines learn how to maintain their planes more efficiently as they do that work – they “go down the learning curve.” – Skateboarders learn to do new things on their boards as they skate. They don’t go into the lab and do R&D – they are learning by doing “Learning by doing” is problem-solving – and so involves trial and error Design • Design a possible solution Build • Develop models prototypes Run • Test model/prototype In real or simulated use environment Analyze • Analyze findings previous step Done Learning by doing can be incredibly cheap for users within their own narrow niche of “doing” Mountain bike innovation • “When I do tricks that require me to take my feet off the bike pedals in mid-air, the pedals often spin, making it hard to put my feet back onto them accurately before landing.” I added a foam ring around the pedal axle near the crank. This adds friction, and prevents the pedals from free-spinning when my feet are off.” Learning by doing can be incredibly cheap for users within their narrow niche of “doing” Example: “I‘m a mountain biker and a human movement scientist working in ergonomics and biomechanics. I used my medical experience to improve my mountain bike. (Consider the cost if that person had not been a biker and had to learn the sport to innovate – or did not have medical training and tools “in stock.”) Table 4.1: Costs of stressed-skin panel innovations developed by users Function performed Time To devel cost solution Framing of openings in panels Structural connection between panels Ventilation of panels on roof Insulated connection between panels Corner connection between panels Installation of HVAC in panels Installation of wiring in panels Connection of panels to roof Add insect repellency to panels Connect panels to foundation Connect panels to frames Development of curved panels Average time and cost for all innovations: Devel cost to wait for mfr Min 0.1 day $20 0.1 30 0.1 32 0.1 41 0.2 60 0.2 60 0.2 79 0.2 80 0.4 123 0.5 160 1.2 377 5.0 1,500 $1,400 $1,400 $28,000 $2,800 $2,800 $2,800 $2,800 $2,800 $70,000 $1,400 $2,800 $28,000 0.5 day $12,367 $153 Exercise: think of your own instance of learning by doing Example: I worked out the quickest route from home to school. 1. Think about the process you used to determine the quickest route. 2. Notice the low incremental cost to you of each experiment. For example, since your trip from home to school was a trip you took “anyway” – you were the “user” of the trip - the cost of each experiment was minimal. (It would cost much more to hire someone to do this experiment who did not have to take that trip “anyway.”) To summarize: Two major factors affect users’ decisions to innovate or buy z z Transaction costs The number of users that want the same product The kinds of innovations that users develop in their “low cost innovation niches” is affected by: z z Sticky information Learning by doing