Product leadership and operating models
I run product as a business. Every initiative names the customer problem, the person it serves, and the metric it should move, before anything is built.
- Where
- iCAD, Komen, SigmaFlow, Ignite
- Leading PMs
- Since 2006
- Model
- Outcome-based teams, monthly ROI gates
- Headline result
- 78% ARR growth; 3 exits
Outcomes
- iCAD: built the ProFound Cloud subscription business from inception to $10M+ ARR in under two years.
- iCAD: 78% ARR growth and 58% U.S. sales growth; renewals rose from 70% to 95% and NPS from 40 to 56.
- Three exits: Ignite Technologies, SigmaFlow (10x revenue growth), and iCAD (>5x acquisition by RadNet).
Transforming product management at iCAD
Where we started
Product management was largely reactive: competing requests from sales, engineering, and customers, with no single view of priorities.
What I changed
- Ownership: PMs owned product lines and their business outcomes, not feature requests, with one prioritized roadmap in place of competing asks.
- Decision forums: a weekly cross-functional product council, a customer advisory group and key opinion leader program that shaped 60%+ of roadmap decisions, and regular portfolio reviews with the executive team and the board.
- Accountability: five outcome-based teams with clear metrics and decision rights, and room for PMs to decide within them.
- Talent: hired and developed senior PMs with the clinical depth to engage radiologists and health-system buyers.
- Customer Success under product: the people closest to adoption and renewals fed the roadmap directly.
The results
ProFound Cloud became iCAD's first on-time product launch in four years. Deployment fell from about 43 days to 3 hours, renewals rose from 70% to 95%, and NPS went from 40 to 56. Product-led expansion contributed to 78% ARR growth and a greater-than-5x acquisition.
Outcome-based teams
At iCAD, I moved product, engineering, and research from functional teams to five outcome-based teams: deployment, adoption, clinical performance, customer success, and innovation. Each owned metrics such as time to deploy, radiologists reading with AI, NPS, and renewals. At Komen, I turned a project-management office into a product-led organization, with teams measured on their outcomes.
An operating cadence that scales
- Annual strategy: themes, major bets, and capacity for 12 to 18 months.
- Quarterly portfolio reviews: where to invest more, change course, or stop.
- Monthly investment gates: explicit continue, change, or stop decisions with owners and dates.
- Continuous discovery: prototypes and pilots with customers before committing to scale.
The same discipline led to hard calls: divesting a business line at iCAD and retiring three products at SigmaFlow to focus investment.
Building product teams
I've led product managers since 2006 across four companies. I hire for judgment and customer curiosity, give each PM ownership of a problem and an outcome rather than a feature list, and evaluate them on results. Several people I've developed went on to lead product teams of their own.
How AI has changed product work
AI has made the cheap parts of product management cheaper: research synthesis, first drafts, and prototypes now take hours instead of weeks. That moves the hard part to judgment: which problem matters, aligning people on the outcome, and earning adoption. I use AI throughout my own work, from research and requirements to clickable prototypes and code, and I set quality bars so AI-assisted work is reviewed, tested, and traceable.
References and related writing
- RadNet to acquire iCAD for $103M (Radiology Business, April 2025)
- iCAD sells its Xoft business to Elekta (MassDevice, October 2023)
- From outputs to outcomes in healthcare product leadership
- AI collapsed the cost of building; judgment is the only moat left