What's wrong with the way the UK funds startup R&D?

Learning from a rejected InnovateUK grant proposal

One perk of researching policy in a co-working space is that you get to meet a wide range of startups on a day-to-day basis and you can see how policy actually works on the ground.

Take InnovateUK’s SMART grants. There’s decent evidence they make a real difference. Startups that receive the grants do more R&D and introduce more new products, according to a robust government commissioned impact evaluation which compared similar successful and unsuccessful applicants.

Funding innovative startups to develop new products is an effective way to increase economic growth, but are InnovateUK’s Smart grants an efficient way to do that? After all, startups applying for funding have to prepare detailed applications which might fail. Some even hire professionals to write their proposals.

In other fields of grant funding, the waste involved is staggering. A viral tweet thread recently summed up the problems. The tweeter cites a study from Nature which found that “For an Australian call, 500 years of research time (= €41 million in salaries) went into the applications for a call with a total value of €226 million.” Similarly, at the tweeter’s own University he estimated that “applying and reviewing for internal funds costs between 8 and 16 million euros for a total budget of less than 50 million.”

For context, the average Innovate UK Smart grant proposal will be 7,000 words in length and typically takes six-to-eight weeks to fill out.

I’ve always been curious if a similar problem exists for InnovateUK grants, so when Lindus Health, a startup I occasionally see around the office, offered to show me their unsuccessful application and feedback from Innovate UK I jumped at the chance to read it.

On paper, the startup in question feels like exactly the type of business that should win funding. Lindus Health is building a platform to increase participation in clinical trials and are working with leading institutions such as Oxford and Imperial College.

They are tackling a major problem in biotech and if successful would not only build a profitable business but also improve productivity across the sector. There’s even a strong case that their business could tackle long-standing health inequalities. Funding for Innovate UK grants is competitive, but if any business could expect to succeed I’d have thought it would be them.

Skimming through their application, I have a few takeaways.

1. Lindus Health pointed out their business, if successful, would improve productivity in clinical research, allowing more research to take place and in turn creating more jobs in the UK. Yet their application was marked down because they failed to consider job losses due to automation. More strikingly, one complained that they failed to “consider the possible negative effect of loss of jobs …. when companies choose [them] over their previous contractor to run trials.”
   
   This strikes me as exactly the wrong way to think about job creation at a time when many businesses are complaining about worker shortages. Suppose Henry Ford was applying for a grant and having listed all the potential economic opportunities cars could create, was marked down for failing to consider job losses in the horse-drawn carriage sector. 

2. The startup’s co-founder Meri Beckwith told me they hired a professional to prepare their application “as there seem to be loads of unwritten rules/expectations, and [we] assumed an agency would know what these were.” However, even with the support of the agency, they appeared to get caught out. “A lot of their feedback was 'more detail needed on x' which seems harsh given we stuck to just within the word limits for each answer. Between the word limits and given that this is inherently speculative, we are not sure how anyone could provide more detail.” Strangely, this was the second time they had applied and the application they prepared themselves was closer to winning funding.

3. Innovative businesses typically experiment, iterate, and pivot until they find a model that works. But the process for applying for grant funding appears to require the startup commits to clear plans over the next five years. This not only wastes time in the application process but can reduce the impact of the grant. As the co-founder put it to me: “Most companies will tell you to discount whatever grant money you receive by 50% to account for your plans very reasonably changing and technology or product features no longer being relevant. Unfortunately, you still need to build them if you want to receive the grant money.”

4. This is very different to the way VCs fund innovation. As Meri put it to me: “VCs essentially assess projects at a very high level of abstraction - are the team hugely ambitious, can they execute, how big could this market be if everything goes right, how well do they know the target customer? Whereas the UKRI process is designed to assess a project at the micro level, but given how fast things change in early stage startups, and the potentially unlimited upside of a lot of innovations, this feels like the wrong approach.”

5. Maybe we should use a different approach altogether. In my paper ‘Unlocking Growth’ and in the Startup Manifesto I helped co-author, we propose copying an idea tried in New Zealand. Instead of relying on the discretion of industry experts, who are often ill-placed to appreciate how an innovation can disrupt or positively transform an industry, applications that meet some minimum standard would be entered into a funding lottery. This would reduce the incentive for startups to over-commit resources on the application process and instead allow them to focus on actually innovating. A lottery model might also end up funding more radical or disruptive ideas. It is worth trying at least. After all, it is strange how un-experimental our approach to funding experiments is.

Stuff I’ve written elsewhere:

Prompted by the response to my thread on the absurdly prescriptive regulations on food high in sugar, salt, and fat, I wrote about the UK’s growing License Raj for the Sunday Telegraph.

You can read it here.

Stuff I’ve read and recommend:

The big idea: could fixing housing fix everything else, too? - Sam Bowman, The Guardian

A Simple Plan to Solve All of America’s Problems - Derek Thompson, The Atlantic

Unblock research bottlenecks with non-profit start-ups - Adam Marblestone et al, Nature Health

Johann Hari’s stolen ideas - Stuart Ritchie, UnHerd