CSIRO needs a standard policy for spinouts
For the last 12 months, I’ve been trying to create a spinout from the CSIRO and I’ve learnt a thing or two about what works and what doesn’t. With this post, I’d like to contribute constructively to the discussion of what a standardised approach to government spinouts should look like and how to maximise their potential for success. This may actually also be useful for spinouts from large enterprises more generally.
I’ll start by telling our story in the context of what is important for startup success, then explain what policies other organisations have created that are being successful at spinning out, and finish with some recommendations.
What is a spinout?
The terminology in the spinning off context is actually quite messy, so let me explain where I stand. Approval to run a mostly self-contained project within CSIRO under a new brand name with a target of creating a new entity, maybe even with an incorporated new company behind it, is not a “spinout” or “spinoff”. As long as the people continue to be employed by CSIRO, all or most funding is provided by the CSIRO, and all or most IP maintains to be owned by the CSIRO, it’s merely the first step towards a spinout. Let’s call this a “spin-in”.
The real test comes when trying to make such an entity independent of the CSIRO and turn it into a company with its own employees. People and who has control over them is at the key of spinning out. What is required is investor money or large paying customers to take this step. Our company was a spin-in for a year before we decided it was time to try to spin out and stand on our own feet.
Speed of decisions
They say some of the key advantages of startups are to be agile, adaptable, unbureaucratic and fast in decision-making. Doing this as a spin-in is hard, really hard, because you have to conform to existing formalised processes while at the same time behave like a startup to your customers and potential investors.
I’m in the ex-NICTA, now Data61 division of the CSIRO and our spinout efforts happened during the conversion period, so don’t take the amount of time it took us as an indication of how long it will typically take for CSIRO to make a decision to allow a spinout to happen. However, the first thing I’d like to point out is that speedy decisions are not a forte of government organisations. It’s not an obstacle - it’s just something you have to calculate with. It also results in CSIRO needing to invest a lot more money into their spin-in projects than would be necessary, because bureaucracy makes the founders and the investors wait.
It’s quite easy to understand: government administrations are typically risk-averse, simply because they have to make sure to have dotted all the i-s and crossed all the t-s before making suggestions about decisions to superiors. Since governments work for the people, all decisions are in the public eye and there is a responsibility to have done the most in-depth due diligence possible.
In the CSIRO, there are at least 4 levels of administration to go through to get a decision made about a spinout: your division commercialisation people, your division head, the Major Transactions Committee (MTC), and the CSIRO CEO. After this, the CSIRO board and the science minister will be informed and may still object, but this is rare. A public announcement is not made until the separation is completely, which in our case depends on investment coming in.
As a CSIRO employee who is trying to spin out, your first step is to work closely with your division commercialisation people. They are your friends, because it is their target to do something with the IP created and will make sure that things move forward. But they are also representing CSIRO’s interest, so you’ll still have to negotiate a sensible position for your startup. Since there is no standard policy for how spinouts work, it will take a while to figure out the conditions.
Once you’ve outlined the conditions, your commercialisation people will need to communicate with legal to make sure everything is kosher, and with your division head to get their buy-in. A bunch of paperwork, business plans, and draft contracts need to be prepared as a submission to the MTC.
Since the MTC only meets about 10 times a year and has many many things to decide, you’ll be lucky to get slotted in within 3 months. Assuming your paperwork is finished, you’re now just waiting to get approval. For a fast-moving startup, 3 months is an eternity to wait - particularly for tech startups and if you’ve lined up investors who are waiting to move forward. Expect to do some more paperwork once you’ve been assessed by the MTC.
The MTC gives a recommendation to the CSIRO CEO, who then signs off on the spinout, giving the task back to your division to finalise the legal paperwork. By this time, you’ll have to update your financial data, pitch deck, etc, because your business approach should already have been market tested and had to pivot. You may need to restart your fund raising, further delaying the actual spinoff date and requiring more investment by CSIRO.
My first recommendation is therefore quite simple: make spinout submissions a priority work item of the MTC that need their own shorter assessment form (the default one is targeted at large contracts) with governance risk being the only relevant risk to assess - all other risks, such as market and financial risks, are up to the founders to manage, not up to CSIRO. Require spinout submissions to be decided within 4 weeks of submission to the MTC (incl. CEO sign-off and legal documents), giving founders a timeline to work towards with investors.
The emotional rollercoaster
Why does the CSIRO need to undertake this much due diligence on spinouts? After all, spinouts are normally small fish at the time that they spin out - they are just startups, and we know that at least 80% of startups fail. So, what is the due diligence for?
It’s about making sure there is a good story to tell about the money that was spent to create the IP and what effort is being taken to get a return for that money for the Australian people. As part of that, it is seen necessary that the CSIRO take as much control of the startup as possible to continue controlling the return on investment. The general position is that the employees who have worked on creating the IP were paid for their efforts, so they own nothing and should be happy about any shareholding and control the CSIRO gives them in the spin out that was created by the organisation. That’s the view of the organisation.
There is one big problem with this position: the spinout is not created by the organisation. It never is. It can’t be, because the organisation is risk-averse and spinouts are risk-bearing. It’s not possible to tell a person in the organisation to create a spinout and take on a more insecure position for less pay. Therefore, it’s never the organisation creating the spinout.
Spinouts are always created by individual people. These people take an enormous amount of risk, they give up their security within the CSIRO for an uncertain and difficult future in a startup that has a high likelihood to fail, and at minimum is going to require much more than a 9 to 5 job at less pay for several years to come. Such a step requires courage and a prospect of being worth the effort.
The founders will full well accept that they were paid to create the IP and that the IP is rightfully owned by the CSIRO. However, at the point of spinning out, they should be treated no different from independent, external people who are trying to turn CSIRO IP into a business.
Instead, founders are often regarded as rebels within the CSIRO: they may leave behind teams less capable of delivering on large consulting contracts, or they rip a hole in a research group that will be hard to replace, or they take away a technology that had started bringing in good industry money through license contracts. Therefore, founders attract the rage of business development, of research leaders, and possibly of their peers who think they are being left behind to fill resulting gaps.
A potential founder, once having shared their intentions, ends up in a very lonely spot with very little support from internal stakeholders and being treated with lots of hostility. The organisation needs to learn that founders are actually helping it achieve its goals and needs to become supportive of founders, not hostile.
CSIRO needs to learn to provide different standard means of how to commercialise IP in a new spinout, needs to learn how to plan the steps to come and give a timeline to work towards, and needs to understand that if it agrees with the way its IP is getting commercialised, it needs to learn to trust the founders to make the most of it.
CSIRO also needs to learn to let go - decisions in commercial businesses are made on a different basis to research businesses so trying to keep control may actually be counter-productive to achieving success, which is all that matters once the spinout has left.
My second recommendation is therefore: start supporting founders by working with them, not against them. Include them in the decisions, e.g. allow them to explain their vision for the spinout in the relevant MTC meeting, have them pitch to their head of division and the head of the CSIRO. Work out a timeline with them for what steps need to be taken, when decisions will be made, a target date for spin out, and what internal investment will be made available until then. In turn, have them help fill any gaps they may leave behind by helping with hiring decisions.
CSIRO as investor or licensor
The role of CSIRO in a spin-out is not actually clear. A startup is a spin-out because it was created by people from within the CSIRO using IP created while they worked at the CSIRO. What the new entity looks like and what CSIRO’s involvement is isn’t actually clear.
Many different models have been used in the past. The most important are:
- license the IP to the new entity exclusively in return for royalties to be started a few years in
- take a shareholding in return for IP assignment
- take a board seat / board observer seat / advisory board seat
In the case of 1, CSIRO becomes merely a licensor of IP to the startup and receives its return on investment from the royalties. It has very little actual influence on the progress of the startup, on decisions and strategy. This is a typical path taken when IP is commercialised by an outside entity.
In the case of 2, CSIRO becomes an investor with a shareholding related to the value the IP is agreed to have in the new venture. As with other investors, this often also means a seat on the board at least for the beginning of the venture while shareholding is high.
These are typical models and all sorts of variations and combinations are possible. Also, CSIRO sometimes does some of the following:
- invest money into the spin-out in return for more shareholding
- provide contractors for a transition period in return for payment or shareholding
- provide ongoing R&D in return for payment or shareholding
- retain an option to take back the IP in case of venture failure
- retain a frequent reporting line, particularly in case of licensing where no board seat is held
Our experience started with NICTA where a shareholding was the preferred option - licensing and royalties was not discussed with us. Data61 has an aim to retain closer ties to their startups and thus requested a board seat in eternity and an option to take back the IP in case our venture failed.
What do other organisations do that create spinouts? One particularly interesting statistic to look at is Herve Lebret’s analysis of equity stakes taken by different US Universities in 2013. His results are informative:
- on average, US Universities take a 10% stake at company creation, reduced to a 5% stake at post–series A (average valuation of: $5M), and a 1-2% stake at exit
- royalties are in the 1-4% range and typically start a few years in
- typically no board seat is requested because there are not enough people to service
As an example, let’s look at Google and Stanford University: research started in 1996, google.com was registered in 1997, Google incorporated with a $100K outside investment in 1998 at which time Stanford received 5% ownership and the founders moved offices out of Stanford and hired a first employee who had been working with them at Stanford. By 1999 they had 8 employees.
My third recommendation is:
—
I don’t know why any scientist or engineer or business person in CSIRO would submit themselves to such madness. Only the most passionate and most determined of employees will go through with this process and nobody comes out unscarred. Maybe I do it because I have a Messiah complex, I’m stubborn, I don’t take “no” for an answer, and I need to prove myself? No matter why, I believe that the CSIRO consist of amazing people who should be supported in their drive to create greatness for Australia, I want to prove that it’s possible to succeed in spinning out and in creating a successful startup, and I want to help to make the creation of spinouts a more common commercialisation strategy for the CSIRO and ultimately for all government bodies everywhere.
So let me suggest a model that may work.
With this, I’d like to encourage CSIRO to start taking the opportunities of spinning out seriously and becoming professional about it rather than muddling through as it currently does with no framework that spinouts go through, no standard pathways, and a lack of expectations for when decisions will be made.
===
Why I left CSIRO and Coviu
- 12 months and no decisions in sight
- I took a salary cut when deciding to do the startup and my capabilities have since not increased - CSIRO is sucking my career prospects out of me
- other startups have come and gone in this time
- startups move fast, CSIRO moves slowly
- support services rule CSIRO and are not there to help the organisation achieve things, but there to check, cross-check and limit what an individual is able to do out of fear they may do something wrong (the biggest difference from NICTA)
- every decision needs at least 4 levels of counter-checking by the boss, BD&C, legal, comms - and all of these people only work 9 to 5 and small projects like a startup are not their priority
- promises to internal projects are never held
- my colleagues at CSIRO don’t understand that I am not here for a cushy job where I can just wait out the time until CSIRO makes a decision to allow us to spin out - my goal is not to have a secure income now, but to make a difference by creating a new company - I may be dumb to want to take on that risk in your eyes, but I’m not here to just sit things out
- spin-ins need to be able to follow different administrative paths to the rest of the organisation
- small contracts are the beginning of something bigger for startups - small contracts are not of interest at all to CSIRO
- SaaS products are alien to CSIRO - small recurring revenue is not how the organisation works - nothing can deal with it: not legal, not accounting, not project management, not BD, not CRM systems
- a spin-in team in CSIRO is completely unprotected from getting ripped apart - if a big contract comes along and people are needed to satisfy it, the organisation simply takes people off other projects, particularly those “internal” projects that don’t make any money yet (i.e. incubated spin-ins)
- CSIRO have no understanding that IP without people is worth nothing - software has no intrinsic value unless the people that have worked on it maintain and further develop it
- in summary:
- CSIRO do not trust their employees
- CSIRO do not value their employees
- for startups, everything stands and falls with the team
- I’ve found it impossible to create a SaaS startup out of CSIRO and after CSIRO having missed three consecutive deadlines I have set for giving us the go-ahead for spinning out (each 6 months apart), the team is falling apart and I have reached the end of my tether
Feedback from Max:
Let’s be real here: creating a spinout is a crisis for a company. Not only is it not in the interest of anyone in the company to achieve this - it’s also a risk that no employee is prepared to stand up for: what if it fails?
Using