Feb 18, 2015 · 30m · mad
Zach Weinberg, Flatiron Health // Using Data to Cure Cancer // Data Driven NYC (FirstMark Capital)
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
At Data Driven NYC, Flatiron Health co-founder Zach Weinberg discusses building an oncology data platform, tackling the technical challenges of unstructured electronic health records, and scaling a healthcare tech startup.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 8.5% of the talking time here. How this is scored →
Matt as informed peer 2.9 Guest teaching 5.1 Guest disagreement 1.7 Matt pushing back 1.3
speaking balance: gold is Matt, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 10:58 Doctors will not change behavior
Zach takes a firm contrarian stance against conventional tech optimism, arguing that startups expecting doctors to change habits or fill out extra fields are mistaken.
Hardest push from Matt ▶ 13:04 Email check on machine learning inconsistencyMatt presses Zach on machine learning by citing an earlier email exchange where he caught Zach making an inconsistent statement about ML's role in their technology stack.
Biggest teaching moment ▶ 8:55 EGFR mutation data extraction gapZach demonstrates deep technical domain expertise by showing how structured EMR fields only reveal EGFR mutations 10% of the time, whereas Flatiron's unstructured extraction reaches 99%.
Matt holds his own ▶ 13:04 Host leverages off-stage prepMatt uses his offline preparation and email correspondence to push back on Zach's public narrative around machine learning.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Matt as informed peer | Guest teaching | Guest disagreement | Matt pushing back | Why |
|---|---|---|---|---|---|---|
| Zach Weinberg's Path from Ad Tech to Flatiron | 2 | 3 | 1 | 2 | Matt opens by asking about Zach's pivot from ad tech to healthcare, lightly teasing him about his story with 'That's not the story I heard'. Zach provides a collaborative overview of their 18-month research phase and thousands of meetings. | |
| Explaining Flatiron Health's Business and Oncology Cloud | 3 | 5 | 1 | 1 | Matt asks about the Oncology Cloud, prompting Zach to clarify that Flatiron is not primarily a data company but a network business. Zach educates on how their EMR and analytics platform cover roughly 20% of US cancer cases. | |
| Network Data Aggregation and Research Use Cases | 3 | 5 | 2 | 1 | Matt seeks clarification on how anonymized data is shared across practices for rare cases. Zach gently reframes, noting that 90% of their focus is actually on the single-player workflow rather than multiplayer data sharing. | |
| Cleaning Messy Medical Records and Unstructured Data | 4 | 6 | 0 | 0 | Matt links the discussion to dirty data formats in medical tech. Zach delivers an detailed explanation of data extraction from faxed, tilted documents and highlights their EGFR mutation extraction jump from 10% to 99%. | |
| Addressing Doctor Behavior and EMR Limitations | 4 | 6 | 3 | 1 | Matt asks why healthcare technology remains so archaic. Zach offers a contrarian view, explaining that doctors will not change their behavior because extra data entry reduces patient volume and income. | |
| Pragmatic Technology Approach and Machine Learning | 7 | 5 | 3 | 5 | Matt references a prior email exchange where he caught Zach making an inaccurate claim about machine learning. Zach acknowledges the callout and clarifies why machine learning is secondary to human-in-the-loop tools for blurry EMR scans. | |
| Series B Investment and Strategic M&A | 4 | 5 | 1 | 2 | Matt probes into Flatiron's $130M Series B round. Zach explains acquiring a larger 55-person company while being a 30-person team, before transitioning to audience questions regarding non-oncology expansions. | |
| Audience Q&A: Encouraging Cleaner Data Entry by Doctors | 0 | 5 | 2 | 0 | The host is silent while an audience member asks if Flatiron encourages doctors to be neater. Zach explains why playing 'Big Brother' to doctors fails and why incentives must align directly with physician workflow. | |
| Audience Q&A: Preparing for Company Inflection and Resource Allocation | 0 | 6 | 2 | 0 | During audience Q&A, Zach addresses company resource splits and reframes a question about FDA clinical trial rules by explaining that Flatiron focuses on the 95% observational market outside FDA registration trials. | |
| Audience Q&A: Handling External Hospital Data and Health Exchanges | 2 | 5 | 2 | 1 | Matt manages time while an audience member asks about health information exchanges. Zach explains that unstructured inpatient faxes captured in outpatient EMRs make formal HIE integrations unnecessary for their model. |
Statements from this episode (22)
Ask for advice rather than pitching during early customer discovery
“Our tip and trick for this was always just to ask for advice. Pretend like you're not actually selling something so people enjoy giving feedback they don't like being sold to.”
Flatiron raised $138M from Google, its largest healthcare deal
“We, we've raised about a hundred and thirty eight million dollars from Google. We're their largest healthcare investment to date.”
Flatiron Health software connects 20% of US cancer cases
“And that network now represents about 20% of all cancer cases in the U.S. So about one in five cancer cases we see.”
US oncology EHR adoption will reach 100% by 2017
“And in oncology that's about 90% or so, and growing. It'll be a hundred percent within the next two years.”
Flatiron Health sees data from under 700,000 cancer patients annually
“We see a little under 700,000.”
Flatiron spends 90% of its time on single-player software value
“Today we've spent probably 90% of our time focused on the single player use case.”
Flatiron Health partnered with NCCN's 25 major US cancer centers
“We announced a partnership with this group called the NCCN, which is the National Comprehensive Cancer Network. It's 25 of the largest cancer centers in the US where if you contribute your data you will be able to get the broader data set for research purposes…”
Flatiron Health spends most of its operational time cleaning dirty data
“We talk about all the fun software, but the reality is we spend most of our time cleaning data.”
Flatiron captures EGFR mutation status in 99% of its dataset
“If you looked for that in a structured field, you'd see it about nine or 10% of the time. In our data set, it's 99%.”
Flatiron tracks a 6,000-person lung cancer cohort monthly
“To do it at scale, we track a 6000 person advanced non-small cell lung cancer cohort every month.”
Doctors will not change workflows within a startup's timeframe
“Doctors are not going to change. And at least not in any sort of timeframe that matters for a startup. Maybe, maybe 1015 years.”
EMRs are built for billing and workflow, not population health
“EMRs, including our own, we're no better, are built for physician workflow and billing. They're not built for population health or informatics or all this”
Machine learning should enhance human experts rather than replace them
“Our take on this is we should be building tools for trained professionals along with a feedback loop to tell them how they're doing. Machine learning and other technologies I think you can apply on top to make those people slightly more efficient.”
Most under-20 person healthcare startups fail to cross the scaling chasm
“There's a hell of a lot of 20 people and under healthcare startups coming out of incubators and whatnot, but most, if you look at the success, and not to, Poo poo startups, but most fail to cross that chasm into, like, a successful healthcare business.”
Flatiron Health spent over half its $130M Series B acquiring an EHR
“We raised a Series B is a hundred and thirty million and we used some of that money you know, more than 50%, let's say, to acquire this electronic health records company that builds a specialized oncology EHR.”
Acquiring an oncology EHR saved three to five years on Flatiron's roadmap
“We probably shaved like three, four, five years off of our roadmap.”
Oncology accounts for about 40% of pharma R&D spending
“So about 40% of all R&D spending from pharma is oncology based.”
Changing physician behavior requires showing direct benefits to individual doctors
“If you're gonna tell a doctor to change something they're doing, you need to tell them why it matters. And I think the key is explaining, hey, if you do this thing that you haven't done, and you weren't trained to do, and you went to school for, like, 15 years…”
Flatiron's business is 60% provider-focused and 40% life sciences
“And right now about 60% of our business is, is provider focused, and then the other 40% Is, ah, research and life sciences focused.”
Flatiron targets data from the 95% of cancer patients outside trials
“So we don't touch that piece of the market. That's about four to five percent of patients will go on a trial. We look at the 95% who are not on a trial.”