Demos are essentially promises: we will create the thing this demo evokes in your imagination
The purpose of a demo is to rally excitement and resources that can get you to reality
Demos are not a sign of progress in and of themselves - often they can actually take work away from progress on the actual thing because in addition to the TRL ~3 system, you need to build a lot of instrumentation to make it useable that will be thrown away and replaced in the final output
In an ideal situation you wouldn’t need demos - the money factory would trust the innovation org enough to give them the resources they need without the demo and understand technical reports enough to understand the progress that has been made
Ideally demos would follow the “apple model” where the demo is just showing off the thing that is already in production
The ability to change direction quickly is tantamount
Anything that makes it harder to change direction should be eliminated. Among other things this includes unchangeable roadmaps and metrics, overhead and restrictions on spending money and hiring, and generally needing to ask permission before taking action.
Research is never linear. Like The Idea Maze, you might think beforehand that you need to zag when you actually need to zig. In the context of innovation orgs, budgets with restricted line items - so much for personnel, so much for equipment, etc. - can severely hamper an organizations ability to execute on
You can say all you want about long term thinking but if you need to constantly go and raise the next quarter’s worth of money, you will be thinking about what you can do in order to raise that money
It’s all about people
There are two ways that people make or break an innovation org.
First, like other high-variance activities all the results are driven by the tails of the distribution. This power law means that the best person for the job is going to be massively more impactful than the second best person and could make the difference between success and failure. As a consequence, innovation orgs with restrictions on who they can work with - whether because of salary caps, credential requirements, or personality tests - are likely shooting themselves in the foot.
Second, the importance of people mean that incentives on those people are critical. If individual incentives aren’t aligned with the organization as a whole, chances of achieving your desired outcome decrease significantly.
Shamelessly stolen from the excellent book Loonshots
In the same way that the world will not beat a path to your door if you build a better mousetrap, if an innovation organization does some amazing work, that work will not automatically impact the world. It requires work from people outside the organization to diffuse it, whether through commercialization, continued public development or other mechanisms. It takes planning and work to make sure this will happen, and too often people focus too much on the organization itself and less on how the organization will hand off the work it has done. Often this involves collaboration before the work is done.
The other part of this principle is that once you have the right people and incentives in place, further external management and intervention generally hurts the innovation org.
Explicitly call out misalignments between the money factory and the innovation org
At the end of the day, the goal of the innovation org is to create things that have never existed before and the goal of its money factory is something else (otherwise it would be an innovation organization itself.) It is uncomfortable, but if you can explicitly call out the misalignments between these two goals, you can come up with a strategy to mitigate the effect of this misalignment. There are several ways to mitigate the misalignment. You can constrain what the innovation org works on - this works when the money factory itself has an extremely broad mandate, like DARPA and the DoD or Bell Labs and AT&T. Alignment between people playing different games can happen on finite time scales so another approach is to explicitly define the expectations and timescales that each party expects before starting.
Research is not linear
You can’t expect regular progress reports from an innovation organization. Einstein published four of is five most impactful papers in a single year, and the fifth paper ten years later.
Innovation orgs are by their very nature trying to do things that have never been done before. You can’t do something that has never been done before by acting in entirely normal ways (otherwise it would have been done!) So in order for an innovation org to work, stakeholders both outside and inside need to be comfortable with actions that are hard to justify or predict a priori. That is, innovation orgs will not work without a level of trust.