How we built our medtech startup without marrying a billionaire

Cyndi Williams
Cyndi Williams

Just over five years ago, Isabella and I met for our bi-weekly coffee and mentoring session. This time she surprised me with a very unexpected topic. She was planning to quit her job and set up a business to make technology to help people with diabetes. She wanted my advice on how to get going, and wondered if I knew anyone who could help her set up a business? Three hours later, I was pretty sure I did.

Two women and a major world health problem to solve

We started meeting in the evenings at Caffe Nero after work. We talked a lot about our principles, what kind of product and company we wanted to create, how we wanted our users to feel, and what kinds of changes we wanted to see in the way they are treated. We also read Founders Dilemmas and talked a lot about our personal drivers for starting a business and discussed all the well-documented things that go wrong in co-founder relationships. The more I learned, the more fascinated I became with the opportunity to make a major dent in a major world health problem and improve hundreds of millions of lives. I knew this was a business that needed to happen and I knew we were the ones to do it.

What does bold look like?

We turned the visioning tools from our past consulting lives on ourselves and did the “what does bold look like?” exercise we had done so many times with retailers, media companies, and banks who aimed to disrupt their industries. Our answer?

When we’ve completed our mission diabetes should no longer exist.

Our users’ lives would be so greatly improved by our product that we could then leverage their knowledge to help identify the root cause of every condition that results in high blood glucose (currently all called “diabetes”). These conditions would then be named, researched, targeted, treated and possibly even cured.

Engineers solving a medical problem

We’re engineers, and we’ve created enough consumer technology to know that the only way to make something users love is to make it with them. 

Early on in this journey, we realised our app would need to be cleared as a regulated medical device.

There’s not much precedent for iterative, agile, consumer-centred regulated medical device development.

We had to break new ground. And we would need a doctor onboard. While most medics we spoke to found it hard to disagree with our user’s pain points, most of them politely declined to get involved because of the lack of clinical evidence for our approach. One high-profile endocrinologist advised that if Quin ever works, it will only be a niche solution because the artificial pancreas will be available within 2-3 years (a favourite headline for at least a decade now). They then ended our meeting with a slightly patronising: “I hope I haven’t discouraged you”.

Having spent 5 years of our lives draining our personal finances to near-zero while hearing constant objections from healthcare professionals, we were not discouraged 🙂  Thankfully along the way we’ve also found a few human-centred, open-minded endocrinologists and researchers who love our engineering approach to solving a medical problem. They believe it will work, and are willing to support us with clinical advisory and evidence collection. Isabella and I are pretty hard-headed, but we admit it matters and we are grateful and proud to have this validation.

Why people don’t use diabetes apps?

Once we were clear on our vision, we incorporated and started our initial user research. There were already 1500 diabetes apps, and most of them were rarely if ever used. We needed to find out why, and learn what would make someone use a diabetes app?

We met up with dozens of people who take insulin in coffee shops around London. We did video calls with many more around the globe. We asked questions, listened, learned, and interpreted their responses in the context of our vision. And we shared our early ideas, concepts and prototypes and asked for their feedback.

Most of what we learned wasn’t a surprise: every person is different; people are tired of being judged for their diabetes outcomes; they don’t have time to log diabetes data nor do they believe logging will help them; they “don’t want to look at maths” (i.e. charts and graphs interpreting data). We also interviewed doctors and healthcare professionals, and learned how far away the deep personal emotions around blood glucose results are hidden from doctors.

Bridging vision and reality

From all of this, we could see our app must be friendly for the heart and for the brain, relieving both the cognitive and psychological load of our users. And that every piece of data entered by them must give them significant and immediate value. Although they didn’t say it in so many words, we discovered that people who take insulin would use a diabetes app if it would make decisions for them by answering daily questions like: “How much insulin should I take for this lunch, and when?”, and “What do I need to do to prepare for this run?”

This opened the door to a whole bunch of other discoveries… Insulin is a treatment that’s difficult to understand, and even when a person understands it, the delayed nature of how it takes effect makes it very difficult to manage. The timing of an insulin injection is just as important as the dosage. And most surprising: even though people who take insulin make hundreds of decisions a day to manage their condition, they’re often not at all aware they’re doing that.

Yes, there’s some trendy machine learning to be done, but first and foremost, we have a massive design challenge on our hands. 

Why doesn’t one of you just marry a billionaire?

Once we had a clear enough idea of what we wanted to build, we needed to hire a team to bring it to life. Hiring a team means raising money – a tough job for an experienced entrepreneur, and an even tougher job if you’ve never done it before and you’re selling something that goes totally against the grain. Literally hundreds of meetings, and dozens of applications. Applications to pitch. Applications to join an accelerator. Applications for industry awards. Applications for grant funding. I have become an expert at describing what we do in 50 characters, 25 words, two sentences, 250 words, 500 words etc.

Like most entrepreneurs, in return for all our work, we got (and still get) a lot of rejection.

Too early. Not enough clinical validation. Not enough customer traction. Too expensive.

Where are the doctors who think this is a good idea? Who is writing the code? (we are both engineers!) How will you ever get enough data? (the joys of “big data” misinformation) Why doesn’t one of you just marry a billionaire? (seriously!)

Fortunately, we also found a few early angel investors who said “yes”. Some who just believed in us, and others who had decades of experience taking insulin, so they totally got the challenge and the need. We’re grateful to them and also to Innovate UK who gave us a grant to progress our current research.

The fun stage

There’s been many highs and lows over the past 5 years, but I have to say for me this is the most fun stage so far. We’ve got 100 people participating in our research programme to co-develop our app and many of them are telling us they’re seeing real improvement in their everyday life. We have an incredible growing team of 12 people – each one of them the real deal in both heart and mind. Our R&D is now moving at a good clip.

We’re building up relationships with some incredible researchers and clinicians and we’ll be putting some more medical meat on our engineering bones over the coming months.

And as always, there’s another chunk of financing to raise. We still have a long way to go to achieve our mission. We’ll be starting clinical studies in the UK and Europe early next year and we should have an initial product ready for launch sometime later in the year. So far we’ve focused on getting regulatory clearance in the UK and Europe (class I CE Mark and soon to be class IIa!), but next year we’ll set up shop in the US, and start the process with the FDA, including a clinical trial. We’ll also start building go-to-market partnerships so we can get out there as quickly as possible once we have clearance. R&D will be full speed ahead throughout, and as our app gets more and more sophisticated we’ll increase our clearance to class IIb, and start looking at clearance in Canada, Australia, Asia Pacific and the Middle East.

I’m fully aware of how blessed I am to have this work to do. Since I started my career over 24 years ago, I’ve never been poorer, I’ve never worked harder, and I’ve never been happier. 




Cyndi Williams
Cyndi Williams