Steven Johnson's wonderful book, Where Good Ideas Come From: The Natural History of Innovation presents a fascinating case for the environments and the mindsets that produce innovation. However, its focus is similar to the focus of many companies that want to 'focus on innovation' - the ignition of an idea (in the book's case, some really big ideas like the theory of evolution). However, innovation has to have a point. It is not the innovation per se, but the fact that the innovation wins out economically, that produces a net gain for anyone. The only innovation that counts is the one that lands.
The focus on innovation within pharma is important, but the diligence and discipline it takes to land the innovation is at least as important. Glivec isn't the drug it is simply because a bright set of researchers had a great idea, but because it was studied and launched appropriately. It only delivers its enormous value to patients because Novartis launched it. The soil that a seed lands in is just as critical to the eventual flowering as the seed itself, as any French winemaker will suggest.
The idea that ideas have an economy seems to sit at odds with the concept of innovation. However, a tagline like 'We Innovate Healthcare' shouldn't mean that a company has cleverer researchers, doing wacky brainstorming, but that the company actually succeeds in putting those great ideas onto the market. Without that nurture, and a productive soil, any innovation would be academic (literally). A company that 'innovates healthcare' is saying that it is better at the disciplines of finding unmet need that others haven't spotted, or delivering/ developing products that others can't, not just that it is discovering better raw products. However, the emphasis on 'clever' biology, mechanisms, science, diagnostics, etc., has not been matched (in most places) with an enhanced discipline of putting innovation onto the market.
The market may be a rather capitalist notion for researchers, and physicians, but it is the only mechanism we (currently) have for making drugs available to prescribers, and thereby to benefit patients. If a drug does not make it to market, it is of no use to any man. The market has an additional requirement - it forces an assessment of value onto an innovation. The economy of the market decides which of the hundreds of ideas (or the tens of new pharmaceuticals) that get launched every year are good ones and which are the losers.
Attrition is a loaded word, suggesting a natural, unpredictable, wearing away of products over time in a pipeline. However that is only half the story. For every molecule that just doesn't work, there is another that doesn't fit into a predefined framework of 'the kind of drug we want to develop'. The 'economy of ideas' operates at every level in a pharma company, from screening, and pre-clinical to marketing. "It seems to do this." "Oh, but we don't want one of those, we want one that does this other thing." The very idea of 'proof of concept' needs a concept to prove, and underlying that is the idea of a disease that needs a better treatment. If that call was wrong, or the concept of 'better' is under-elaborated, then the study is just a fishing trip - useful academically, perhaps, but we may never know.
Innovative companies are the ones that have ways to understand what could be valuable on the market, and mechanisms to simulate that economy internally. All of which makes it odd that so many companies focus on taking the opportunity out of opportunity by doing the same market research, buying into the same landscape reports and following precedent in the clinical development process - just like every other company that they compete with.
The equation of innovation with ideas is incomplete. Innovation is what happens when ideas land and win.