Same same, but different
Choosing what matters and defining better health outcomes in digital health.
“We’re going to cure all disease by 2100 with AI” - Zuckerberg Chan initiative, 2025
I remember reading this in an article a few months ago and almost choking on my sandwich. All disease? Cured?
Ambitious? Yes.
Important? Yes.
Any understanding of the actual things you need to do beyond “tech” and “more investment” to achieve this? Nope.
Accountability for such an outrageous claim? Still no.
A tension
I’ve been wrestling with this over the last few months with my clients.
On one hand, the digital health industry needs to show evidence of real health outcomes that matter to people beyond accuracy and in silico benchmarks.
As the industry matures (kicking and screaming, and not uniformly) across the Gartner hype curve, there is a pressure to stand out. Moving beyond operational time savings or better patient engagement, to the harder work of answering the So What? question.
How does this efficiency gain lead to something important for patients, clinicians or the system? What measurable outcomes does this track to? This is a welcome progression. It’s what matters.
On the other hand, health outcomes take time to achieve. Plus there are many confounding factors. Your digital tool, super game-changer AI model is being implemented into a complex adaptive system that is unique in its messiness and just by observing it, it changes beneath you. An intricate weave of human pathophysiology, human behaviour, change management, socio-technical determinants. Policy, procurement, ethics, social acceptance.
Stuff you can’t vibe code. And isn’t really what investors or public perception has the patience for. Not when we have a “once in a generation opportunity” or the “horse has bolted”.
So what happens is that we hedge, ambitiously leading to the type of statements we see at the start, emblazened on billboards and in shiny conferences and polished decks. You get a strategic elasticity: whatever ends up happening, you can just draw a line back to this vague mission.
Ok, well what’s wrong with that? It’s a BHAG! (Big, hairy audacious goal)
Ambitious goals stretch us. Moonshots are what get creative juices flowing.
The problem with these sweeping vague assertions, are that there’s no definition of success, nor milestones. There’s no accountability by design. So this becomes just more noise that washes over us. Its just the same vanilla flavour. They also show a deep lack of understanding of the barriers to curing ‘all disease’, as if their vision or product is the ‘missing piece’. More on that later.
Crucially, for those building it affects internal product and commercial decision making. I’ve seen this across many teams and companies manifesting as a lack of clarity that confuses and attenuates day to day momentum.
Teams can’t align.
The end points are vague and open to interpretation. People struggle to align to one direction. Efforts are attenuated. Commercial sees it one way, product sees it another way, clinical folk see it yet another way. Circular conversations. Messaging misaligned with product roadmaps, a lack of credible evidence. Clients churn.
It’s hard to say no to anything.
Everything becomes a priority, so nothing is.
“The goal is to cure all disease” : good luck with the product needs prioritization session with that as the north star.
A different way
After wrestling with this at startup, scale up and established company level over the years, what I can say is that there isn’t a ‘hack’ for this. And fancy OKRs don’t automatically solve it.
It is about taking a moment to get intentional and specific. Not just ‘strategically’, but doing the work to make sure that translates to tactics and day to day decision making. If you are a leader, welcome that scrutiny from your teams.
Remember, anyone can make claims. Do you think anyone is going to judge the Zuckerberg Chan initiative for not reaching their moonshot goal of curing all disease by 2100? It’s just spin, and there will be more spin to cover the previous spin. You see right through it.
You only win in the long term if you actually deliver and prove that.
The word ‘Trust’ is banded around like a mythical concept at conferences.
“How do we build trust?”, the thought leader says in a packed room as they gaze thoughtfully into the distance. Everyone nods in admiration. “How do we create true impact”. The audience is star struck.
I look down at my conference buzzword bingo card. I have won.
But how do we do those things really? Well we can start with proof for what we claim.
The right claims: Grounded and specific, not vague.
We do know this from the work around SMART goals, but often we get caught up in the race, what competitors are saying and investor pressure. So we expand the goal instead of narrowing.
Teams yearn for focus. Clearer, faster decision making. What’s helped me is asking : Does this help us to identify specific things we SHOULDN’T be working on? Can we stress test this and refine our decision criteria?
When I’ve worked with companies this point is a real a-ha moment that forces a deeper set of questions. What do we stand for here? Better quality conversations ensue.
Specificity helps prioritisation. And you need to make some pretty robust choices on what you get specific on.
The right proof: Robust, defendable.
To do this well, you need a few questions to consider.
Most teams stop after they’ve identified some ‘metrics’. Things get interesting if you question further. Can you actually move this number? What other dependencies in the system affect this? How deeply do you know the context? What is your hypothesis that shows how this tool or model changes a decision or behaviour that leads to your clinical outcome? This is the starting block to get your methods, analysis and assumptions to stand up to scientific rigour.
This approach grounds you towards what really matters in the clinical setting. It helps sift signal from noise and put you on solid ground.
I’ve seen teams visibly relax at this point. A renewed sense of clarity and somewhere real to direct the energy and purpose: a lever point. What was previously a big, ambitious and fuzzy goal, is now a tangible outcome that matters, with a specific roadmap and clear milestones to get there.
And finally that guy at the conference can see how you create impact and build trust.
See the bigger picture
We’ve put the latest shiny tech on a pedestal since we invented flints in caves. I’m paraphrasing a bit from a video with Meredith Whittaker, President of Signal who calls this out really well:
Innovation is not just using the latest version of technology.
We think that if people could just use GenAI, we’ll finally solve “disease X” or “access to care”. We’ve had the treatment to curing one of the deadliest diseases in human history, Tuberculosis, for over 80 years. It is not an innovation/tech problem. It’s an us problem.
Look again at the quote at the beginning of this piece:
“We’re going to cure all disease by 2100 with AI” - Zuckerberg Chan initiative, 2025
If you buy into the barrier to us curing “all disease” is a tech capability one, I am compelled to counter this quote with a few from John Green, author of Everything is Tuberculosis:
“TB follows the path of injustice that we humans blaze for it… Where are the drugs? The drugs are where the disease is not. Where is the disease? It is where the drugs are not.”
“What does that say about us? What does it say about the world we’ve built? And the world we might build instead..”
Don’t blindly accept the fallacy that tech’s ability to improve access and solve humanity’s problems is pre-ordained. It could, but the jury’s still out. It depends on us and the choices we make. We need to get intentional, not reactively automate because “this is a once in a generation opportunity”.
Product teams are measured by their ability to ship fast. That’s great, but we risk going fast in the wrong direction and ending up at a dead end, with people harmed at scale, places we don’t like at all.
If you are building or implementing a technology, dig deeper under what you see at the surface. Think about the wider causative determinants. Where does your tool help or make things worse? What choices can you make? What are the second and third order effects? What are the unintended consequences? How do you help make peoples’ lives better? How do you, yes you reading this as a person working in healthtech, make choices in your work to not leave people behind?
Do we really have the space to have these conversations? I think we do and can be doing much more. Good software design, risk identification, intended benefits, clinical evaluation plans: these are all established opportunities to ask these questions (and answer them).
The next time you find yourself in endless alignment meetings and circular conversations, or rolling your eyes at a bland statement about ‘improving the health of a billion people’ in a company all hands, get people to turn those into tangible, provable and well defined outcomes that can make things clearer.
If you’re wrestling with this at your organisation and need someone to help you navigate, I’ve worked on this with some of the biggest health tech companies in Europe. Get in touch at hello@shubs.me
Interesting stuff in Global and digital health
I’ve been quiet on here the last few months. Here’s what I’ve been up to:
Freelance clinical leadership advisory with digital health scale ups: Building robust ways of monitoring LLMs. Not just about correctness, but about real value in a consultation. I’ve really enjoyed getting technical again with data scientists, PMs, devs.
Building Healthtech hero with my great friend Lorenzo Espinosa: we launched first cohort last year and delivered deep dive workshops with Doctolib for their product onsite to 80 PMs, UXR and design folk.
Delivering a keynote at Doctolib’s Product Offsite to >250 PMs, PMMs, clinicians on what value in healthcare really means, and how to deliver it. A huge highlight, and so great to interact with people who really care about clinician wellbeing and patient outcomes.
🎥 Recorded the first in studio video podcast episode in Nairobi with Patty Mecheal, CEOof healthenabled.
🎧And check out the latest podcast episode with Javier Elkin about how the International Committee of the Red Cross built out their digital innovation roadmap, and what LLMs need to be able to do in complex humanitarian settings.
📣 News in Global digital health and tech in low resource settings
Horizon 1000 : Gates and OpenAI have launched this initiative to support Primary Care clinics and integrate clinical decision support across facilities in Rwanda, then Kenya. I’ll be interested to see the outcomes this is looking at and how they work to achieve them.
Global Digital Health Events:
🇨🇭🌍 Geneva Digital Health day: We’ll be hosting a Fail Festival. Are you implementing in low resource settings? Is your work on Implementome?
I’ll be hosting a track at HIMSS Copenhagen in May - get in touch if you are going!
Past events and insights
🇮🇳 India AI summit : I wasn’t there, but got a great set of takeaways from Nick Martin
🇰🇪 🌍 December 2025 saw the very awesome Global Digital Health Forum in Nairobi. I collected a series of hot takes from the companies, policy folk and funders who were there.
Hot takes videos : some spicy, some mild!
A short quick takes about what innovators need from policy makers from awesome folks I shared a panel with. Insights from India, Indonesia, Lebanon and around Africa.
That’s all I’ve got for now! Look how much awesomeness you got through. Like, the opposite of doom scrolling. :)