What a GovTech Feature Actually Returns

When a CEO asks me to justify a feature, I never start with the feature. I start with the cost it removes. In public-sector work, the most expensive thing an organization owns is the call center — every confused citizen who picks up the phone is a measurable, recurring cost. So I frame feature ROI around deflection: does this screen let someone complete a task alone who otherwise would have called, queued, or abandoned and reapplied? That's the return. A well-designed status tracker, a plain-language eligibility check, a 'where is my application' page — these aren't UX niceties, they're load reducers. The discipline I bring is refusing to spread investment evenly. I'd rather ship one self-service path that's genuinely complete and trustworthy than five half-features that each still funnel people to a human. The other half of ROI is risk avoided: an accessible, mature-feeling service doesn't generate complaints, press, or legal exposure. For a CEO, that's a real line item even when it never shows up as a number. My job is to point investment at the few features that quietly retire cost — and to say no to the rest, loudly.