A question we get often from cancer center directors is “what should we fund next?” When you have an inbox full of pitches saying “our tool is essential”, it could be difficult to choose what gets funded and what doesn’t.
One oncology service line lead shared, “We get 200+ genuine proposals annually, each claiming to be essential. Our budget funds maybe 5-7 major initiatives.” We spoke with another cancer center executive who said they have roughly 100 projects in flight at any given moment.
That’s the honest scale of what a health system is juggling when a pitch lands in their inbox. A start-up is competing against three other vendors claiming to solve the problem they solve and competing against 100 internal priorities.
So, as a cancer center / service line, what should you fund next?
Build the foundation first
A cancer center CEO I spoke with put it this way: “If your EMR is terrible, don’t buy AI tools. Fix the EMR first. If your data is fragmented, don’t buy analytics platforms. Foundation before flash.”
One industry leader told me that when he was hired to create an innovative plan to redesign a cancer center, the first page of his memo was all about parking.
There are three main reasons someone shows up at a cancer center: they’re a patient, they’re visiting a patient, or they’re a donor. In the building he was working on, all three had to fight through a backed-up valet, confusing parking, and a walk past a designated smoking area just to reach the front door. None of it was patient centered, and none of it required a single piece of new technology to fix.
EMR stability, core data infrastructure, cybersecurity and basic operations tools first. An ambient documentation tool bolted onto a broken EMR will fail. And, it will take the next three pitches down with it when leadership decides “we tried innovation and it didn’t work.”
Have an honest conversation around ROI
Return on investment is usually simple: (Benefit minus Cost) divided by Cost. It gets messy in healthcare because benefits show up over years, not immediately, some of the value is financial and some isn’t, and every dollar spent has an opportunity cost.
Cancer adds a layer on top of that. Attribution is hard (roughly 40% of cancer patients have at least one other chronic condition. Adults with cancer aged 65+ can run from 60-90% with a comorbidity). And, cancer costs are also rising well over 10% a year, so a vendor claiming a 20% savings needs to be asked: 20% against what baseline, and how is it measured?
I asked a healthcare system’s CEO how they think about budget and ROI. They separate strategic value from economic value, and evaluate both across three different time horizons at once: this fiscal year, the next three years, and the long term beyond that.
The vendors that get funded speak to the buyer’s shortest and longest horizon in the same conversation. They make their ROI case concrete, and make it something they can track and be held to.
Clearly lay out the costs of the partnership
Every technology purchase has an obvious cost listed in the pitch deck and a hidden one: the cost of making the new system actually work in practice.
For example, let’s take an ambient documentation deployment across one clinic. The listed cost runs around $250K in year one: license, implementation, training. The integration cost, the custom EMR connection, the data flow setup and testing, ongoing maintenance, IT support time, could run another $200K in that same year. That’s nearly double the number in the vendor’s slide deck.
Founders who don’t budget for this get blindsided by their own health system champion, who now has to explain a cost overrun they didn’t cause.
You’ve heard executives share no margin, no mission. If a health system doesn’t retain some margin, it stops existing. A start-up partnership has to save money or generate revenue in a way the finance team can rely on.
What technology tools make sense to look at first
Tier 0: Core operations tools (the foundation)
Scheduling system that works. Access is the front door, and in oncology days-to-first-appointment is a competitive metric.
Billing/RCM systems capturing revenue. Before automating anything, make sure the baseline revenue cycle isn’t leaking.
Basic patient portal. Table stakes for everything below. You can’t do ePROs, navigation, or digital intake if patients have no digital front door.
Oncology data aggregation and analytics. The “build the data warehouse first” mandate. A single cancer patient’s record is scattered across the EMR, lab, pathology, genomics vendor, and outside systems; unifying it is the purchase that makes every purchase below work.
Tier 1: High-ROI administrative automation
Ambient AI documentation. Oncology’s long, complex, high-stakes visits make this a necessary tool.
Prior authorization automation. The thing standing between diagnosis and first treatment is usually authorization, not medicine.
Infusion scheduling optimization. The unglamorous math that fills infusion chairs and smooths the mid-day peak. Becker’s reports capacity management as a top 2026 cancer-center priority.
Call center AI / chatbots. Deflects the routine half of call volume (refills, directions, scheduling).
Tier 2: Patient experience and navigation
Patient navigation programs. Keeps patients from falling through the cracks between diagnosis, treatment, and survivorship.
Telemedicine infrastructure / remote patient monitoring. ePRO-based remote monitoring is associated with fewer ER visits and hospitalizations.
Financial navigation / counseling. Cancer is the most expensive diagnosis most families face; unaddressed financial toxicity becomes abandoned treatment and bad debt.
Tier 3: Clinical innovation and research
Clinical trial matching AI. Accrual is both a mission metric and a revenue line, and screening is the bottleneck.
Clinical decision support systems. Keeps treatment decisions on-pathway and defensible, and gives the service line visibility into off-pathway variance.
Genomic data integration. Makes precision oncology operational: structured molecular results, molecular tumor board routing, variant-to-therapy matching.
Predictive analytics (complications, readmissions). Flags the patients likely to end up in the ED before they do, so navigation and triage resources go where they matter.
Tier 4: Next-frontier bets
Novel AI applications. Treatment response prediction and drug development assistance.
Virtual/augmented reality. Patient education, procedural anxiety, rehabilitation.
Multi-cancer early detection. Blood tests that screen for dozens of cancers at once.
Hospital at home models. Delivering chemotherapy and acute care in the patient’s home.
If you are 1) working on something tackling any of the above or 2) looking for solutions in this space, please reach out.
Ten rules I keep coming back to
ROI first, innovation second. Cool technology that doesn’t deliver value is a waste. Technology that delivers value is treasure.
Foundation before flash. Fix the EMR, the data infrastructure and security before you seek to implement AI moonshots.
Pilot fast, scale smart. Test quickly in controlled pilots. Scale only after you’ve proven value.
Buy platforms, not point solutions. An integrated platform beats a pile of best-of-breed tools that all need custom integration.
Budget for integration. The license is half the real cost. Integration and ongoing support often equal or exceed the sticker price.
Hire talent, don’t just buy tools. Tools only work with people who can implement, manage and optimize them.
Measure relentlessly. What gets measured gets managed. No data means no learning and no improvement.
Kill quickly. A failed initiative should die within 90 days. Sunk cost fallacy is what actually kills innovation budgets.
Communicate constantly.
Make innovation sustainable. Embed it in the culture, build self-funding mechanisms, and design capability that outlasts whoever is currently in the leadership seat.
You cannot do everything right now to bring innovation to the front lines of cancer care, but you can decide, with real discipline, which one or two you fund first, model honestly over five years, and build on a foundation solid enough to hold whatever you stack on top of it next year.
Comedy!
I like sayings. I think they’re unique mementos. Somebody wanted to sound wise, without doing any of the actual work of being correct. And it worked, because here we are years later still hand-stitching them on pillows.
“You can’t have your cake and eat it too.” Yes you can. That’s the whole point of cake. Nobody has ever once had a cake and not eaten it. Oh no, I’m fine. I’ll just look at this one.
“The ball’s in your court.” Which was first said by a guy who has watched 0 seconds of any sport.
I understand the shape of it, until you remember that in every sport where that phrase would apply, both teams play on the same court.
“Let’s cross that bridge when we get there” is just “I don’t want to deal with your problem,” but it sounds thoughtful instead of lazy. Which, respect. Also, it assumes a bridge is going to be there. What if it’s a canyon with nothing around? Now you’re just stuck mid-metaphor with a problem and no infrastructure.
Or when a co-worker walks past you on the way out of the office late at night and hits you with the “you burning the midnight oil?” Talk about a saying that’s been running on fumes for 200 years.
You have to “take all of these with a grain of salt.” No more than a grain. As if the correct dosage of skepticism is a single, microscopic crystal. Somebody should’ve workshopped that. What about “Take it with a healthy amount of doubt”? No, no, we’re doing salt. One unit should prove the point.
My favorite saying to picture is “killing two birds with one stone.” Which is efficient, if you were already out there killing birds. Fun to imagine the guy, the rock, the two birds, and a group of people nearby applauding his time management.
And the worst saying passed down as advice is to “bite the bullet.” Soldiers used to bite down on a literal bullet during surgery, before anesthesia. We took a man’s absolute last resort and turned it into a productivity tip for sending an e-mail to your boss to ask for PTO.
Most of these sayings aren’t wisdom. They’re a way to end a conversation with your mouth instead of your brain. Nobody built these to be true. They built them to be short enough to fit on one side of the pillow.
Ask
One of our portfolio companies is hiring a Senior Sales Lead
Ideal candidate has direct experience selling into community oncology (or an adjacent specialty) and working knowledge of oncology EMRs and practice operations.
If you are interested in learning more, or know someone great who could be a fit, let me know.
I’m opening up sponsorship on the Oncology Ventures newsletter for the first time
We are not making a dollar on this
Every cent goes toward two things I care about - increasing donations to cancer charities and building a real community around oncology innovation
50% goes to one of our favorite cancer charities
50% funds an intimate dinner in a city of your choice, where you will sit alongside our portfolio company CEOs, healthcare executives and our LPs
Plus, you will get a feature from me in this Substack that now goes out to 17,500 readers
Reach out if you are interested in learning more
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.