back to article 'Disruptive innovation' is nonsense? Not ALWAYS, actually

If the theory of disruptive innovation is wrong then why do companies act as if it's true? Proof is rather in the pudding after all..... The latest topical shouting match in my corner of the economics-meets-tech world is an essay by Jill Lepore in the New Yorker. All this talk of “disruptive innovation” is just that – talk – …

  1. James 51 Silver badge

    Google would be sucking more out of us like a giant leech with the data that would be feed back from the cars to 'monitor and improve services.' Are we going to have to pay them for the privliage as well?

    1. greifpad

      Agreed, i always assumed that future automated cars would by default gather information from their surroundings and feed it back to a central system for various purposes e.g. updating maps, real time traffic reports, spying on your driving habits etc. etc.

      1. Ben Bonsall

        Surely an autonomous car would be spying on it's own driving habits?

        1. James 51 Silver badge

          "Please enter the google id of every passenger to start thr car."

  2. Pete 2 Silver badge

    Too much law

    The biggest barrier that disruptive technologies have to face today (it was different in the past: even up to 20 years ago), is litigation.

    There are a lot - the vast majority - of companies that make stuff and profit from what is here, now, today. They value the ability to make this stuff so much that they've managed to persuade the legislature of the whole dam' world to grant them patent protection of every stoooopid little idea, component, design flourish, technological variant and process they could possibly push through - and the bar is set pretty low.

    As a consequence, any new improvements on existing products that small companies are foolish enough to try and promote will get stomped into the ground by queues of lawyers focused on protecting the vested interests of their behemoth paymasters. Irrespective of what benefits these innovative ideas and products might bring.

    That means it's getting harder and harder - and taking longer and longer - to "disrupt" an existing incumbent. The only areas open for innovation aren't to disrupt existing, originality-squashing, products but to branch out into new areas. Areas that have no stultifying pre-existing IP and where it will take the legal smother-pillow 20 years to catch up and kill off your novel ideas. But by then, you'll be the industry dinosaur: desperately fending off all these new, warm-blooded, innovators and trying to kill them off, in the same way others attempted to do-for you a generation ago.

    1. Gordon 10 Silver badge
      Thumb Down

      Re: Too much law

      I don't buy that argument - patent madness is mostly limited to the states and it will only work for so long. Outside of the states new entrants will start to eat their lunch.

      And by the by I don't believe truely disruptive products are defendable by patents - as generally thr y cone with new uses that are orthogonal to the existing ones.

      It took less than 5 years to smash Nokia's core business to Flinders and I don't believe Nokia's senior managers are more or less stupid than those of the average globo corp (even if Elop was/wasn't a Trojan horse). That's disruption in action. And they had patents out the wazoo.

      1. Anonymous Coward
        Anonymous Coward

        Re: Too much law

        Companies don't patent a complete product, they patent a tiny part. Even patenting something used on a PC then applies to a jetpack because the patent was allowed to be excessively broad.

        You mention Nokia but the smartphone market has had a massive stream of lawsuits for anyone deemed to be getting too much of a market share and a lot of these patents are all pre-smartphone era. It is usually only a defensive patent folio that forms any sort of protection.

        And to dismiss the whole of the USA as a market you will not approach due to patent issues limits your business plan by quite a margin.

        1. Gordon 10 Silver badge

          Re: Too much law


          I wasn't dismissing the US as a market - merely stating that for a truly disruptive product it doesn't have to be your first market.

      2. TheOtherHobbes

        Re: Too much law

        >I don't believe Nokia's senior managers are more or less stupid than those of the average globo corp

        Nokia were already failing with insane internal competition, project sprawl, and bureaucracy. Core engineering was always world class. Manudjment was increasingly Harvard-led and dysfunctional.

        But they could have recovered with good management - which Elop clearly wasn't.

        As the original article points out, a lot of supposedly disrupted and failing corps go on to have long and happy lives.

        I think we'll see Nokia back in personal comms once the MS agreement has run its course - and they'll still be better at it than MS is.

    2. big_D Silver badge

      Re: Too much law

      I agree with Pete up to a point. The problem is, most oil companies, for example, have invested heavily in patenting tech that would threaten their revenue stream - alternative energy generation, electric vehicles etc.

      That means it is often also difficult to start up in new industries, because borderline competitors have already worried about what would happen there and have their patent blockers up and their lobbyists on the Hill...

      Outside America it is a little better, but still hard.

    3. hugo tyson

      Re: Too much law

      I agree it's too much law, but it's not only/mainly patent law. It's regulatory capture, intentional or not.

      For example motorcycles with roofs, seatbelts and roll cage (and weather protection) still requiring a helmet; taxis versus Uber; new "cars" which should be light and efficient being required to be as over-equipped with airbags and electronics and ABS, ASC,... as old fashioned heavy metal + oil ones. Construction laws making new building materials, power schemes, heating systems... infeasible.

  3. meh1010


    I really would suggest reading Tim Harford's "Adapt" - he does a great job covering this as well as how (and way) only a minority of larger organisation occasionally manage to produce and exploit their own disruptive tech.

    1. BlueGreen

      Re: Homework @Tim Worstall

      To add to meh1010's recommendations, may I also suggest The Effective Executive by Peter Drucker. Old but not outdated. I found it very thought provoking, and it's pleasantly short. It specifically covers obsoleting yourself before your competitors do.

      Can't comment on this article but I look forward to violently disagreeing with you again in the near future.

  4. spider from mars

    I think it would be prudent to wait and see if Google's car succeeds or fails before you use it as the central plank of your thesis..

    1. ratfox Silver badge

      There wouldn't be many interesting analytical articles if they were only allowed to predict what happened in the past.

      1. spider from mars

        The article's not about Google's cars, though; I would be perfectly happy with an article predicting whether they will succeed or not.

        The problem is that the author is trying to use his prediction as evidence, to make a wider point about innovation.

        1. I ain't Spartacus Gold badge

          The problem is that the author is trying to use his prediction as evidence, to make a wider point about innovation.

          No he's not. He's simply using it as an example. The supposed reluctance of Detroit to cooperate with Google, as they naturally don't fancy seeing sales of cars plummet, if everyone goes to a different ownership model. Whether this is true in this case doesn't matter, they may eventually cooperate, sales may not drop, all that matters is the discussion of the motivations.

          I was having this discussion yesterday. We have a niche product that is close to unique, but quite expensive. It's technically more complex than our competitors, in order to achieve a smaller size. That, and the fact that we don't manufacture in bulk, makes us more expensive.

          However several companies have recently attempted to copy us, and mostly failed. It took a lot of testing to get this thing working, some of them even buy from us, so they can keep the customers but sell our product when nothing else will fit.

          Should we make it cheaper? We've tested with a cheaper major part, which will give us an even smaller product for little loss in quality. But it will cost tens of thousands to re-tool, re-design, re-certify etc. Or we could sell as a new product into a different market.

          However if we do either of those, guess what happens to the sale of our more costly product?

          The new smaller one won't cover all use-cases, so we'll be left with loads of engineering and design costs, 2 incompatible product lines and therefore extra warehousing and stock to pay for. And sales costs to push the new product in a new market we don't currently address.

          So our alternatives are to do nothing. Move to the new product and hope we can increase sales to cover our losses (maybe, but impossible to tell), or maybe dump the old product and lose the high end sales that the new one can't address. Sadly these are of course the most profitable ones...

          This would be an innovative and disruptive product in the new market (that we don't currently address), although it's only an incremental improvement in ours. But if some new company came along, with none of our baggage and existing costs, it would also be disruptive to us.

          The reason others tried to copy us of course, is that we were disruptive when entering our current market, because we could do something that none of them could. And this product was originally designed for yet a third market, where we used to exclusively operate, and with only a slight re-design got into our current market.

    2. DaLo

      Something can be disruptive even if it doesn't eventually succeed. A really good company would notice the disruptive threat and react quickly and effectively, using their experience and knowledge which then may make the newcomer redundant. It's when an established company doesn't notice or dismisses the threat, of which there are many examples, that the newcomer can really take off.

      So, if Ford produce a driverless car that works as well as Google's but is based more around single ownership/family ownership and a regular car design (retrofitting to their existing chassis for instance) then that may well be a lot more appealing than an expensive bubble car from Google that has shared public ownership (and there is never one around in the rain or on a Friday night). Google's car might then fail but it will still have been disruptive and caused a seismic shift in industry thinking and speed of development.

    3. James Micallef Silver badge

      Google was a small startup search engine maker that disrupted the then-current search engine behemoths.

      I'm not sure it counts as a small startup robocar maker disrupting the current car behemoths, as even if it's robocar division is relatively small as an individual entity it is still backed by a mammoth company with very deep pockets. Detroit could crush an independent robocar startup, they can't do that to Google.

  5. DropBear Silver badge

    I have no idea what Google's actual plans for their driverless cars are, but I really don't see what the ability to drive autonomously has to do with the form of ownership of a car. Driverless cabs and a personal car that can take me home all by itself after I had a few beers are two equally legitimate but totally different things - I just happen to care not at all for the former, while I would be quite interested in the latter.

    1. I ain't Spartacus Gold badge


      If I want a car to commute to work, but don't use it during the day while I'm there, I could send it away to someone else who works nights, wants to shop during the day, or walks to the office where they're a sales rep, and drive to meetings.

      This would mean I wouldn't have to buy a car, but could rent/hire it only when needed. Saving me money, and I could avoid paying parking.

      To make it scale, I guess it would only really work, if the cars could drive themselves between users. Otherwise more people would use existing car-pooling schemes.

      There are good reasons that business model might not work, but if it did, the car manufacturers would be very sad to see their sales drop.

      1. James Micallef Silver badge

        There's a reason that every major city in the world has 2 rush-hours a day, and that is that the vast majority of vehicle users are using them at 2 narrow time windows in the mornings and evenings. This means that the peak numbers of humans needing to move at the sme time will not be that far different from teh peak numbers of humans who own a vehicle.

        To really reduce number of vehicles on the road, commuters will need to accept to share vehicles with others going (approximately) the same way, as currently the vast majority of rush hour traffic is 1 person driving a minimum 4-seater car. Just get an average 2.5 people per vehicle will halve traffic, greatly reduce commute times and leave commuters with a lot more free time (if they use the extra time themselves) or productivity (if they use teh extra time at work) or even both, since they are now free to follow pther activities than driving during the actual commute.

  6. dotdavid

    You know I'm not entirely sure that a driverless car rental model will have *that* much of an impact on car sales. I would imagine that new car sales (that the manufacturers are paid directly for) are only a small fraction of total car sales so remember it's not as high a sales target as it looks. With driverless cars being used almost constantly they will be wearing out and needing to be replaced with new ones on quite a regular basis, and if a good proportion of the people who would normally buy used cars instead rented driverless ones that might equate to extra sales.

    1. James 51 Silver badge

      The point about cars laying idle all day is a good one but at rush hour everyone one will want a car there and then and to bring them to their own home and wait twenty minutes while they nip into the shop. It would reduce sales but not by huge margin.

      1. TheOtherHobbes

        But the rush hour means that driverless cars will still be lying around idle all day. They'll just be lying around at the GoogleDepot[tm] or on grass verges and in random parking spaces.

        They won't be outside people's homes, unless that particular piece of tarmac is a convenient parking space.

        This may not sit well with punters who want a car now.

        Even with Google's world-class stat skills and prediction models, there will still be people who want a car and have to wait for one. Those wait times will either make the service or kill it.

      2. Richard Gadsden

        Funnily enough, this is already a solved problem

        Ever heard of public transport?

        A lot more people want to travel during the peaks than in the middle of the day or in the evenings. But people don't like travelling by train if the last train is 7pm (they get worried about being trapped) so train operators need to run services until 11pm or so.

        For commuter trains, they tend to run units in multiple in the peaks, and in single outside the peaks, so half the carriages sit idle outside the peaks. Most of those trains go and sit in the depot from 10am to 3/4pm and after 7pm.

        It's difficult to shorten long-distance, high-speed services for technical reasons, so they tend to sell off the spare tickets cheap - that's where all those cheap advance tickets come from.

        For the cars, they will need enough cars to move all peak-time travellers at once, which is about 90% of the number of commuter cars currently on the roads. The other 10% represents people who are off-sick, on holiday, working from home, etc on a particular day. At the moment, those cars are sitting parked outside that person's house; with automated shared-cars, they just won't have one come and pick them up.

        But all the other cars, the ones that aren't used in the peaks in the first place, all of those disappear. Every retired person who still has a car, that goes - they rarely travel in the peak, so when they need a car, there will be plenty available.

        Non-peak car travel will be cheaper, because the cars will be available. Peak-time commuting will be quite expensive.

        Shifting to a renting model makes the prices more comparable to public transport, which might result in some commuters shifting to public transport. At the moment, if you need a car at all, you pretty much have to own one. Once you own one, using it for a journey is usually cheaper/better than using public transport. If you don't own a car, you're paying the capital costs amortised across each journey, the same as public transport - so they can compete on a journey-by-journey basis, and public transport is usually cheaper per mile.

        If you live close enough to a city centre (I do) then you can choose not to own a car, use one of the on-line short-term rentals (car club, zipcar, etc) when you want to make a journey for which public transport isn't good, and use public transport when that is available. I expect that style will be much more widespread when you can live in a suburb and a car will turn up on your doorstep five minutes after you request one - especially as you can get the car to shuttle you to a train station, allowing a single station to cover a wider area.

  7. Any mouse Cow turd

    You've already covered some of this...

    Tim, you've already described how the disruptive innovation can only really get a foothold if the existing regulation (or lack of) allows. -

    So for the likes of Google to change the market with driverless cars the uphill struggle against current regulations will be immense and the incumbents will battle all the way. Your analogy to the automobiles vs trains of the late 1800's is a good one in that there wasn't much/any regulation and probably only applied to horse and carriage.

  8. Yet Another Anonymous coward Silver badge

    Monopolies innovate

    The transistor and most of the semiconductor industry was invented by ATT almost every other part of a computers was invented by IBM. Small agile startups invent the Yo app

    The real problem with innovation now is that there aren't massive companies with total domination of the industry for 50-70 years

  9. Christopher E. Stith

    Sorry, Bell Labs...

    Sorry, Bell Labs...

    It seems the laser, transistor, Unix, C, C++, Hamming codes, CCDs, wireless LANs, fiber optics, and 32-bit microprocessors were completely non-disruptive technologies. So much for your seven Nobel prizes and two Turing Awards.

    1. Anonymous Coward
      Anonymous Coward

      Re: Sorry, Bell Labs...

      > It seems the laser, transistor, Unix, C, C++, Hamming codes, CCDs, wireless LANs, fiber optics, and 32-bit microprocessors were completely non-disruptive technologies. So much for your seven Nobel prizes and two Turing Awards.

      Interestingly, although these are innovative, they are not disruptive.

      I suggest that perhaps Tim didn't make that distinction very well in his article although I do agree with the main thesis that newcomers can only disrupt incumbants.

      1. Pete 2 Silver badge

        Re: Sorry, Bell Labs...

        > Interestingly, although these are innovative, they are not disruptive.

        I would suggest that the transistor was the single most disruptive technology of the 20th century. Above nuclear power/weapons, antibiotics, air travel or even fiat money

        The degree to which it opened up new possibilities and changed our way of thinking was unsurpassed.

        1. Anonymous Coward
          Anonymous Coward

          Re: Sorry, Bell Labs...

          > I would suggest that the transistor was the single most disruptive technology of the 20th century.

          In what way? Sure, it enabled whole industries to exist and made a lot of money for a lot of people.

          But which existing incumbants did it affect other than valve manufacturers?

          I would differentiate disruptive from transformative in this respect. Also, I would differentiate society (customers) from industry (manufacturers).

          1. Pete 2 Silver badge

            Re: Sorry, Bell Labs...

            > In what way? [ was the development of the transistor disruptive ]

            Before transistors: all electronics was bulky, fragile (even metal-enveloped valves), consumed a lot of power, gave off a lot of heat, required high voltages, was expensive to build and assemble.

            After transistors: electronics became miniaturised, portable, low power, low voltage, highly reliable, extremely densely packed (which allowed much higher operating frequencies). It also drove all the other miniaturising processes (such as MEMS sensors) and nano-technology that we are _starting_ to exploit, now.

            These changes moved "electronics" from being one or two large brown boxes in a household and nowhere to be seen in an office, factory or shop to being everywhere and facilitating every single thing we do, being involved in the manufacture of every single thing we touch and becoming ubiquitous in the daily lives of almost every person on the planet. Even the 2Bn people who have no mains electricity rely on mobile phones to improve their lives.

            1. Richard Gadsden

              Re: Sorry, Bell Labs...

              This is a highly-specialised meaning of disruptive, as meaning "there is already an industry making X; a disruptive technology appears, this makes Y; people making X go out of business." Transistors put valve makers out of business, but other than that, all the other new technologies were either greenfield (ie didn't put anyone out of business, but did something really new) or innovative rather than disruptive.

              Take PCs: People using wordprocessors didn't replace people using typewriters; the same people switched from using one to the other. That's classic innovation rather than disruption. Email: disrupted the Post Office and Royal Mail, but didn't result in email-using-lawyers replacing letter-using-lawyers (so innovation, ie process-improvement, not disruption, ie business replacement).

              Digital cameras were disruptive to film-makers, but innovative to camera-makers (Canon and Nikon were dominant in SLRs and still are in DSLRs).

              1. Anonymous Coward
                Anonymous Coward

                Re: Sorry, Bell Labs...

                > This is a highly-specialised meaning of disruptive

                Indeed, but wasn't the whole point of the original article about how entrenched incumbants cannot innovate radical, disruptive changes because it would destroy their cosy status quo? Newcomers don't have anything to lose.

                I would argue that digital cameras were immensely disruptive especially to the likes of Kodak, a manufacturer that was big into the core technology but lacked the managerial will to move with the times, a rather excellent example of the genre.

              2. DocJames

                Re: Sorry, Bell Labs...

                People using wordprocessors didn't replace people using typewriters

                Have you ever seen a picture of a typing pool?

                Seriously, wordprocessors have hugely disrupted the world of work. And this isn't always good - now everyone taps out many of their own documents, rather than dictating them then getting on with some other work. Same with presentations - my parents used to get the graphics dept to do all their slides. Now they make their own slides on powerpoint, and take about twice as long to do so.

  10. James 36


    Personally I think the secondhand patent market should be killed, patents for inventor only, they should not be saleable items.

    On driverless cars as a committed petrol head I like the idea of people who see cars simply as a means of transport having an option that gives them the convenience of a car without the hassles of ownership.

    i want to own a car as I like cars but if I could have access to a driverless car for commuting or social trips, or even family hoildays (including family camping which might fall outside of the use case) then I could have a wet ware driven car for fun,

    This would make me happy, though making me happy probably won't lead to global success.

  11. Irony Deficient

    One of the three great management books of all time

    Tim, how could Il Principe not have made your top three?

  12. User McUser

    in 1895 the car was clearly inferior to the train, the dominant transport technology, for actually getting somewhere. But the car offered the opportunity to get anywhere, which is what was so disruptive.

    Cars were indeed disruptive, but not to trains. Trains still exist and are widely used to perform the same basic functions (hauling people and/or freight) as they always have despite the many improvements to highway systems and vehicles. They were, however, very disruptive to animal-powered conveyances which are now relegated to novelty status at best.

  13. Richie 1

    The Google car business model

    We already have an industry devoted to renting cars to take you on just the journey you need: they're called taxis.

    The Google disruption to this industry is that autonomous vehicles make for cheaper taxis, since you don't need to pay a driver.

  14. Michael Wojcik Silver badge

    Tim's wrong about the PC

    IBM most certainly did consider the impact of the PC on their existing business. In particular, they always believed the PC would complement their minicomputer (System 3x and Series/1) and mainframe (3x0) lines, in three ways: serving as cheap1 terminals, putting prettier and friendlier front ends on mainframe applications, and encouraging data processing in general among middle management.2

    What eventually caught IBM by surprise with the PC was its extremely rapid growth in computing power and storage, so that it could start eating into the minicomputer and mainframe markets. That was driven largely by commodification: if private users hadn't latched onto PCs for gaming, and later for online entertainment and other personal uses, they wouldn't have been picked up by Asian manufacturers as a commodity and prices wouldn't have been driven down so fast, nor would we have had such an economy of scale. That, in turn, made it profitable for near-monopoly players like Intel to pour huge amounts of capital into R&D and improve PC capabilities so quickly.

    But IBM played the PC innovation game too, with a host of evolutionary advancements in various areas, from low-level technologies (SOI and copper interconnects for CPUs, for example) to products like PowerPC that put pressure on Intel and others to keep improving.

    More importantly, for the sake of Tim's claim, the PC wasn't disruptive. IBM famously stumbled in the late '80s and early '90s, but it recovered without discarding its core hardware lines of business. Minis - specifically the AS/400 aka i Series aka System i - were a cash cow for many years; the RS/6000 UNIX business grew solidly and remains intact; and mainframes are still quite profitable. While IBM did diversify by making its service business a much larger component, the one major LoB it did discard was those supposedly disruptive PCs.

    1Because their costs - both monetary and in terms of cognitive load, desk space, etc - would be amortized by their other uses, unlike dedicated 3270s and the like.

    2Once your middle managers get used to crunching local-office budgets in a PC spreadsheet, they're more likely to be interested in things like OLAP, which require more horsepower in your back-office mainframe systems. At least that's the theory. The same argument can be made for things like office document and retrieval - and indeed we saw the various waves of interest in document imaging and later, as more documents originated in digital formats, business content management systems.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019