$76 billion a year from a tableful of products
[Via asymco]

During the calendar year 2010 Apple spent nearly $2 billion in R&D. That is a significant increase from $714 million in 2006. However, as a percent of sales, R&D spending has decreased. Sales have grown more rapidly than resources hired to develop the products (or to sell them).

In Q4 2005 Operational Expenses (costs which are not tied directly to units of production–sometimes called fixed costs) were 14.2% of sales. In the last quarter of 2010, the ratio was 9.2%. Sales and administrative expenses (which include advertising, promotion and overhead) were 7.1% and R&D (which includes all engineering, testing) were 2.2%. As percent of sales both reached new lows.

[More]

I mentioned this last week but this is a nice graph to see trends.

This remark was pretty cool to think about:

The efficiency with which Apple creates sales is legendary. There can be many explanations for this but the most telling evidence of causality I can find is the small number of products in the portfolio. Tim Cook stated that given the sales value, there is more concentration of product at Apple than at any other company except perhaps an oil company. All the products Apple sells can fit on one average sized kitchen table and they generated $76 billion in sales last year.

I’ve written about Apple’s approaches, as well as those of other 21st Century companies. One of the  key aspects is their efficiency – they produce a huge amount of innovative work that belies the small size of their research groups. Pixar is another example.

I worked for a company that accomplished this – Immunex. We were able to compete with much larger companies by focussing our research efforts very tightly. Not that we only worked on a few things. We worked on a lot of them.

But we only allowed a very few to pass through to real development. We did this by having a very open and transparent vetting process for projects. We examined every research project 3 times a year in a process that could be attended by everyone.

What this did is make it very hard to carry on projects purely for political reasons. In many companies, a powerful sponsor could take possession of a project and push it through, resulting in something that dies on the market; in these companies, politcal pull can bemore important than actual innovation.

Because of the social aspect of our vetting, it became much harder to say that a project would continue “Because I said so” when everyone could see that another project held better value. It decreased the ability of politicians to get a project approved.

And, since everyone’s views were heard, people could understand why a decision was made – it was made in public.

We worked hard to kill projects or put them on the backburner. But it was all done in public. And then we allowed every scientist to spend a percentage of their time working on a project – any project – that they wanted to. This not only allowed people to continue for a time even on killed projects, hoping to rejuvenate them, but allowed really creative ideas to be examined.

But, after a certain time, each of these were vetted in public. If they did not pass the public test, they were scrapped.

We developed a lot of innovative ways to foster creativity but the most important was making the discussions and decisions open to anyone, make the important decisions as early as possible and make them public.

I don’t know if Apple does the latter but it seems likely they do the former, as this from asymco suggests:

Most observers of technology are not aware of the pace of its development. It’s natural to assume that most R&D costs are in the product creation, or early phases of development. Coming up with something new must be hard. But that’s not actually true. Most R&D work is routine polishing of products and coordination late in the development cycle. “Productization” is far more resource intensive than “concepting”.

This is absolutely true – coming up with ideas is easy; making them happen is hard.

It stands to reason that making go/no-go decisions early in the pipeline is a lot less expensive than making stop-ship decisions prior to launch.

I have no specific evidence that this is the case, but I guess Apple conceives of plenty of concepts, but chooses to move forward to develop and market very few. Most companies don’t have the ability to decide early and proceed with costly R&D and marketing in order to find out whether products will “work” in the marketplace. The proliferation of flawed products is a big cause of the inefficiency of product development.

Look at so many companies. They put out lots of products hoping the market will decide which is best. They want consumers to do the hard work. How did the Kin see the light of day? Many companies just can not kill a project that has emotional and political connections. So they let it slump on through, hoping that the market sees potential.

That is why, a year after the iPad, we have 100 different copy cats, while in the years before the Ipad there was just nothing even close. Most companies have no real idea of what is successful to they copy other’s success.

But Apple, like Immunex, repeats innovation time and again because it has developed a process of killing things that do not work and, if they can, killing them before they progress very far. They may not always succeed – they have had failures – but they usually know why it failed.

I worked for a company that came up with innovative solutions for years –again and again – so I know that what Apple is doing is not only working but is reproducible.

That is what separates a 21st Century Company from a 20th Century one.