Thinking about innovation and business models

I’ve always maintained that people who “think opensource” work on useful things, solve problems, create value; they don’t focus on the business model at the outset but instead concentrate on the value they create.

In Peter Drucker’s words, “people make shoes, not money”. Make something that is worth while and people will pay you for it. Figure out what shoes you’re good at making and then make them well. You will make money as a result.

Knowing in advance how you’re going to make money from snake oil may sound like you have a business model; what you have is snake oil. And that’s the problem you need to concentrate on first, the fact that you’re not creating anything of value.

And sometimes the process of calculating and measuring benefits can come in the way. Many years ago, when I worked for Burroughs Corporation, I learnt this the hard way. This was the early 1980s, and software/services was just emerging as a business. Until then, all the margin was in hardware, so we ‘shifted tin”. We gave away the software and the services in order to sell the hardware. Then, as the cost of human capital rose, and investable capital became scarce, this equation began to shift. It became more and more important to understand the true cost of software projects before starting them.

So we instituted something called the Phase Review Process, borrowed from the US Navy if I remember correctly, and implemented it within the firm. Every project had to undergo a phase review at inception and then at each phase.

Which was all fine and dandy. Unless you were just about to start a project that would cost a total of £25,000 inclusive of everything. Which was less than the lowest possible total cost of the phase review process. But I was lucky, my management understood this issue, and it was mandated that projects had to exceed £100,000 in total planned cost before they needed to be put through the Phase Review Process.

Why am I writing all this? Well, some years ago I remember reading about something called the polypill; the newspaper articles referred to this paper which had been published in the BMJ in 2003.

The principle was simple. Six tried and tested medications to be combined into one pill that could cut potentially reduce cardiovascular disease by 80%.

When I first read the articles, I was intrigued. But I didn’t know much about the drugs involved. I knew nothing about statins, other than some vague notion that they were wonder drugs that combated high cholesterol with some wonder side effects. I knew even less about ACE inhibitors and beta-blockers, though I may have come across the beta-blockers as something to do with performance enhancement. Folic acid was something pregnant women took; and diuretics meant you had plumbing problems.

Aspirin I knew about, although I had no idea it could be obtained in cardio doses.

But that was in 2003. Since then, as many of you will know, I have had reason to get to know this particular cocktail of pharmacology quite intimately. Nevertheless, I’d forgotten all about the polypill.

Until a few weeks ago, when I read this on the BBC web site. The polypill could become reality in five years’ time, it said. And then I remembered what i’d read all those years ago, when they said … that the polypill could become reality in five years’ time.

And that made me think. Slowly. Very slowly. And my thoughts went a little like this:

One, cardiovascular disease is the single biggest cause of death facing humans.

Two, people had come up with a cheap and effective way of reducing the risk of cardiovascular disease by 80%.

Three, this had happened six or seven years ago.

Four, with a little bit of luck and a following wind, we may see something happen in five years.

Of course I’m oversimplifying, but I don’t believe I’m exaggerating. A strange world we live in.

I’m not by nature a conspiracy theorist. I believe man landed on the moon nearly forty years ago. I don’t believe in little green men or UFOs. Neither do I believe that Big Oil makes sure that substitutes for gasoline never surface.

But here is what I believe. I believe there is some evidence that the polypill does not exist today because it’s hard to make money from it.

Why? Because the ingredients in the polypill are all out of patent, all “generic”. Because the way drugs are trialled, it’s prohibitively expensive to bring a new drug to market unless you have some monopoly rents to come, patents to exploit and exhaust.

So it is possible that the cost of trialling a cocktail of generic drugs exceeds the potential income from selling the cocktail. And so no polypill.

No mention of the number of lives potentially saved and minor stuff like that.

Now I take statins, beta blockers, ACE inhibitors, diuretics, blood thinners and anti coagulants daily. You could say I have an amateur interest in all this. A passion, even, given that the medication has worked wonders on my heart and on my life expectancy.

This is not meant to be a diatribe against doctors or the medical profession or even the pharmaceutical industry: they have all treated me really well, and I owe them a debt of gratitude.

What I am trying to do is to point out that sometimes we hold up innovation by concentrating on the wrong thing at the start. And sometimes it’s because of the anchors and frames of the way we do things.

So I was thinking. Opensource people solve generic problems. Is there a way to opensource the trials of generic drugs, to change the mechanics and dynamics of drug trials for generics? Is there a way to adopt the opensource principle of “privatising losses and socialising gains”, the exact opposite of what happened during the credit crunch?

I wonder.


19 thoughts on “Thinking about innovation and business models”

  1. Hmmm – I must do some digging into my archives.

    I met a chap last year at an exhibition in London who was discussing a cross-vendor “project” being chaired by a chap at Roche that was looking into using Open Source methods and principles for managing pharmaceutical trials and the data captured.

    There are some FOSS projects IIRC that are designed to provide the “tools” to help with the technicalities of trial management but that isn’t quite what you were thinking I believe.

  2. I think there is. It would be a case of generics companies creating new combination of drugs, doing a lower cost toxicology study and then essentially farming it out to doctors and patients to try the new combination. The doctors and patients would then provide regular data. The number of patients that would try the variant is likely to be very high making recruiting easier.

    It is really a change in methodology around what is needed to trial a drug. The concept of a double blind test would but the very size of the trial is likely to reduce placebo effects. The data would be harder to analyse as more would need to be controlled for (diet changes, other drugs being taken) but again the potential size of the trial would make this easier as large number of variants would exist.

    I think once a drug is approved for use, then the methodology for trial the drug for new diseases/delivery etc. should not be the same as trialling a wholly new drug.

  3. The Cochrane Collaboration does something similar to what you suggest:

    “The Cochrane Collaboration is an international not-for-profit and independent organization, dedicated to making up-to-date, accurate information about the effects of healthcare readily available worldwide. It produces and disseminates systematic reviews of healthcare interventions and promotes the search for evidence in the form of clinical trials and other studies of interventions.”

    “Those who prepare the reviews are mostly healthcare professionals who volunteer to work in one of the many Cochrane Review Groups, with editorial teams overseeing the preparation and maintenance of the reviews, as well as application of the rigorous quality standards for which Cochrane Reviews have become known.”


  4. JP:

    In order that biotech/ pharma can adopt a more Open Source approach to trials, more than just the system of economic rents needs looking into. The entire academic publishing industry, which props up many a career and many more jobs, is biased towards positivity. Have you ever come across a trial which was published and actually had disproved the hypothesis? Those experiments and ensuing papers (assuming they are written) never see the light of day. Which means that only a select portion of science is really up for scrutiny. Contrast this with open source code. How easy is it to hide crap code? Not very, once many eager brains are crawling all over it. Right? So the problem goes much further back before the pharma industry, doctors and patients get involved.

    I clicked through on Mark Littlewood’s link too and I have to say that health will remain in the hands of pharma industry and the doctors until prospective patients decide it is in their interest not to get ill and unproductive in the first place! Alas the P-word gets no attention from investors for the case of X-people-saved-from-death is only statistically made, and never provable (Can we run control groups on long-term mortality studies? I am not sure how we might do that except statistically). Ergo the approaches we see now.

    JP, I am trying my best to get to ICA this evening with a bumper headache but I am very keen to meet you in person. Aspirin, here I come!

  5. Thanks for your comments everyone, I will try and respond tomorrow, fairly stacked today. Sorry.

  6. Simon’s comment above is important, but can perhaps be cast in terms of the implicit financial motivations that drive this system.
    The level of testing that is required to produce a new drug ‘implies’ a financial value to human life (this is an unpleasant thought, but true). And it does so in two ways. One in terms of how much we should pay to avoid unintended consequences of the drug (risk avoidance through multiple layers of testing) and then (if its a drug that saves live’s) through how much we are willing to pay for the drug at the end of the process. Ideally the setting of these costs should ensure that (in an ideal market) that capital is deployed in a way that provides the most benefit.

    Of course it is not. And the relative implicit values will be completely different in the case of the blended generic drugs.

    All this is by way of saying that thinking about ‘business model’ and thinking about solving important problems SHOULD be quite closely aligned. They are not, and particularly so in the case of the Pharma industry, where our tolerance of risk is set so microscopically low that innovation cannot happen.

    It is a political issue.

  7. I second Glenn’s comment tha part of the problem is we are so risk averse that even an opensource model for trials could be a stretch unless we are prepared to do a couple of things –

    1) Use modeling tools

    After all the generics are (or should be) well understood individually so the issue is really can a polypill (in your example) deliver the same effect as taking a cocktail of pills each with its own specific delivery mechanism (coated/uncoated) etc.

    2) Introduce far greater scope for personal choice.

    There is a strong argument for personal decision making here. Clearly at one extreme if my life expectancy sans a particular drug therapy is be less than the time it would take to complete a trial, then I’m considerably more likely to opt for this join the trial as an open source contributor even if the trial data was imcomplete.

    Can’t resist observing that here we are talking about an opensource model where we are the software! It’s getting late.

  8. JP, that is good insight, and I fear that you may be right about the cause of the “delay” in this superdrug.

    Yet I still think the best answer remains within the pure scientific process. One of the commenters above mentioned the excellent work of the Cochrane Group, and I particularly recommend the writing of Dr. Ben Goldacre in his incessant argument for the same issues that concern you, including the need to register all trials even (especially) where they do not show a positive outcome.

  9. Thanks JP,

    This made me think and helped me understand some stuff.

    But it took me in the other direction to the question you pose at the end. Made me wonder if as well as the opensource model hinting at a way through pharmacology challenges, there wasn’t the risk of pharmacology problems hinting at where opensource could run into problems too.

    Just as the ‘for-profit’ tech industry has hit problems – Moore’s Law meaning leading to businesses being unable to support the products they provide (see Nick Jones of Gartner isn’t there a risk that there are some tasks / problems where the open source approach just requires too much effort to justify doing it without ‘snake oil-like’ commercial engines to drive value realisation?

    Haven’t got my head round that properly, and am unable to express clearly – but there’s a reverse to think about, too. That’s my point.

  10. People often like to “fondle their hammer” and skip over the value creation question and this limits or diminishes the innovative process. Too often, in business settings, we are driven by the need to drive revenue and profit for the “things” we do or have rather than start our thought process from the knowledge of what the “customer” wants or needs. We also spend too much time trying to protect our ideas and assets and believe this is the stuff the differentiates our competitive advantage. I look forward to the day when more and more people “think opensource” because this is when we will see some dramatic breakthrough ideas.

  11. On a tangent to the main here: Phase Review Process is patented by Robert G. Cooper, and the excellent book PACE, “Product and Cycle Time Excellence” gives a good introduction. I did a (small) bit of research on medical trials and CEBP and guess what, there’s an application. Medical trials need LOTS of people to partake in late stage trial, because people don’t take their meds on schedule, don’t record their intake properly, and don’t take them EXACTLY when they are supposed to. Using CEBP decreases the amount of people and time you need to run the trial because the trial is run more accurately.

  12. Alvin Toffler in one of his books .. was it Megatrends ?
    suggested that the way forward is to rubbish the idea of patents.

    Which is what you are driving at.

    My take..
    > Glory be to the intellectual pirates
    > Thumbs down to the businesses that are rent-seeking

    I need get my head around how to achieve this utopia
    in a gentlemanly manner..

    Any comments / ideas ?

Let me know what you think