Toyota makes cars. Google makes connections. Toyota takes a set of raw materials and transforms them into a product people buy. Google takes the information stored in the internet and makes it accessible to you and me.

Google, Facebook, and many other internet giants are intangible marketplace providers. They provide the cyberspace market where customers for bits and bytes link with providers of those bits and bytes. The customers do not expect to pay directly to access the bits and bytes, and once accessed, may or may not pursue the service or goods the bits and bytes represent. Considerable ingenuity is required to monetise the market, or platform, as it is commonly described. Google does it by piggy-backing advertisements onto those transactions. Anyone with access to the internet is either a bits and bytes fabricator, or a customer, or both. So anyone with something to sell is a potential Google advertiser. This is why it is so wealthy. Many other would be platform providers work on creating the platform first, and monetising it later. Which is why investing in internet start-ups is a risky business.

For over a decade, we’ve used lean thinking to redesign the way healthcare organisations deliver care. Many lean concepts are useful to us, but lean has it’s restrictions.

How Google works is written by Eric Schmidt and Jonathan Rosenberg, the professional managers hired by the founders of Google to manage the company as it expanded.

Whilst Google occasionally ventures into the world of things, essentially it is a pure knowledge-work organisation. Its basic resource is the knowledge its founders, and their employees, have of applied mathematics and computer programming. For elegance and sheer density of ideas per page, How Google Works does not compare with Taiichi Ohno’s Toyota production system; beyond large scale production. But How Google works is all about the management of a company of knowledge workers, a subject in my experience, lean does not do enough to address.

The Australian healthcare classification system groups all the individual diagnoses a large public hospital expects to deal with into around one thousand different individual, resource related, diagnostic related groups. A public hospital of any size has to be able to recognise and deal with most of the 1,000 clinical groups, alone, or in combination, and modified by a host of person-specific factors. It also has to be able to respond twenty four hours a day, seven days a week, 52 weeks a year.

Yes, there are a range of conditions that present frequently. But public hospitals cannot just say they have to keep a whole lot of stuff to manage small-volume problems. That is not very cost-efficient. Let’s just stop doing the complicated or the unusual and stick with the high volume work. That may be possible if you are a for-profit business; not in public healthcare.

How does the public healthcare system do it? By employing a large number of trained doctors, nurses, scientists-and, within limits, giving them the autonomy they need to practice their knowledge based trades. It is this autonomous, knowledge based work that is the stuff of healthcare.

I am not a fan of management or business writers. One exception is Peter Drucker, who had great ideas. Back in the 1970s, Drucker was already talking about knowledge work and the knowledge worker. He said for the knowledge worker, the crucial question is ‘What is the task?’ For the process worker, the question is ‘How the work should be done’. Both groups can be creative, but the knowledge of the knowledge worker is what that worker uses to define the task. The customer, the patient, presents with a problem. It is up to the healthcare worker to reconceptualise that problem as a healthcare problem, decide what that problem is, and design and execute an appropriate plan of action. For the process worker, the task, the design, is given; the challenge is finding the best way to make the design work.

Drucker added a further distinction: between pure knowledge work, and hands-on work. He used the example of a surgeon, who does the pure knowledge work of examining, assessing, choosing and explaining, and the hands-on work of cutting and caring. Most healthcare work involves a combination of pure knowledge and more structured doing. There may be one best way to do doing work, but the one best way for knowledge work is only visible in retrospect. For Drucker, when the balance tilts towards hands-on doing, the worker is a technologist rather than a pure knowledge worker.

I am not sure about the ultimate validity of the distinction but I recognise as a healthcare redesigner, most of my work is at the junction of the pure knowledge and the technical. I have to leave the pure knowledge stuff to the knowledge worker. I am a psychiatrist by background, and even though I am a qualified medical practitioner, I have to leave the heart stuff to the cardiologist. Only she understands her pure knowledge work issues, so both as a redesigner and an occasional patient, I, and the managers who work with her, have to trust she and her colleagues adequately police the quality of their own knowledge work.

As a redesigner, my challenge is how to get the cardiologist, and all the other knowledge workers, to accept whilst I don’t want to interfere with her clinical decision making, I would like her to start her ward rounds on time, and do the quick work of confirming discharge decisions first, rather than leaving them until last. I want to redesign and improve how tasks get done, but the tasks are still her’s to identify and design.

Healthcare workers are used to a degree of independence. Knowledge work cannot be formulaic. When I am sick, I want someone who treats me, not the average patient. I learnt I could not get these independent souls to do things by telling them “upstairs are worried about your turnover. The boss has told me to come and sort things out”. Healthcare workers are notoriously effective at passive resistance. If they do not want to do something, somehow, it just does not get done. I can see their point, having myself been subject to attempts by inexperienced project officers to impose poorly thought through solutions dreamt up by ambitious senior executive trying to gain favour by managing up.

It took us some time to work out how to differentiate gaining the permission of knowledge workers for their work to be part of a programme of redesign, from getting a programme of work authorised by the management team. Authorisation and permission are both important, but require quite different tactics.