3.2.3: Case Study - The Two Masters

The Citizen Portal: Reporting to Two Bosses at Once

A federal modernization initiative called the Civic Access Program had been handed 450 million dollars to bring outdated, disconnected government systems up to modern standards, so citizens could interact with public services without the maze of separate logins and paper forms they'd been stuck with for years. From outside the agency, the mandate sounded almost simple. Inside it, Naomi Patel, the product lead assigned to build the program's flagship shared citizen portal, was discovering exactly how complicated simple could get.

The portal was supposed to serve two departments at once: the Bureau of Retirement Benefits, which processed millions of routine transactions a year, mostly people checking payment status or updating an address, and the Bureau of Veterans Services, whose caseload ran through complex, multi-step claims that often needed a caseworker involved at every stage. Naomi's team reported, on paper, to both bureaus equally. In practice, that meant reporting to Deputy Commissioner Diane Okafor from Retirement Benefits and Assistant Director Marcus Webb from Veterans Services, two executives with entirely different ideas about what the portal should actually be, and neither one inclined to defer to the other.

The disagreement came to a head in a design review that was supposed to settle the homepage layout. Diane wanted the landing page built around fast, self-service transactions front and center, arguing that most users just needed to check a payment or update a record and shouldn't have to click past anything else to do it. "My citizens don't want a case management system," she said. "They want to look something up and get on with their day." Marcus pushed back hard. "My veterans are navigating claims that can take months and touch four different offices," he said. "If their case status isn't the first thing they see, we've built a portal that works for her bureau and fails mine." Naomi's team sat in the middle of it, unable to build either version without the other executive treating it as a loss, and the compromise design that eventually shipped satisfied neither: a homepage split awkwardly between quick self-service tiles and a case status widget that didn't have enough room to actually show what Marcus's users needed to see.

A comfortable assumption had been circulating at the agency's senior levels that this kind of conflict didn't really require the kind of formal governance a private company might set up, since the program was funded through a fixed congressional allocation rather than driven by profit and shareholders. The agency's CIO said as much when Naomi first raised the stalemate. "We're not a corporation chasing quarterly numbers," he told her. "We don't need a governance board sitting over two departments arguing about a homepage." It turned out to be exactly backwards. A private company with a bad quarter has revenue elsewhere to absorb the cost of a poorly coordinated launch. A public agency spending a fixed appropriation has no such cushion, and a stalled or embarrassing rollout burns public trust in a way that doesn't come back just because next year's budget arrives on schedule. Without a governance body positioned above both bureaus, there was no natural place for Diane and Marcus's disagreement to actually land. It simply kept recurring, sprint after sprint, at whatever level of the org chart happened to be in the room that week.

The stalemate became impossible to ignore once the compromise homepage actually shipped to a pilot group and the Veterans Services call center saw a 30 percent spike in complaints about not being able to find case status information, the exact failure Marcus had predicted in the design review nobody had fully resolved. That spike reached a congressional staffer working on program oversight, and the possibility of an unflattering hearing did what months of internal debate hadn't: it got the agency's deputy secretary to authorize a Digital Modernization Steering Committee sitting above both bureaus, with the authority to make a binding call whenever departmental priorities collided on a shared deliverable. Naomi built a RACI matrix for the portal that named a single accountable owner for every major decision instead of leaving accountability split between Diane and Marcus by default, and the steering committee used it to settle the homepage fight directly: case status would lead for logged-in users with active cases, and self-service tiles would lead for everyone else, a decision neither executive had proposed but both could live with once someone above them had actually made the call.

The portal relaunched with the resolved design, and complaint volume at both bureaus dropped back to normal within a month. But the steering committee itself existed only through the deputy secretary's interim authorization, tied to her tenure rather than written into the program's permanent charter, and a change in agency leadership was already expected before the next fiscal year. Naomi raised the gap at the committee's own retrospective. "This worked because someone above Diane and Marcus was finally willing to make a binding call," she said. "I don't know yet whether the next reorganization keeps that authority in place, or whether the next shared deliverable between two departments starts this exact same fight over again with nobody positioned to end it."

Discussion Questions \(\PageIndex{1}\)

1. Naomi's team spent months absorbing the conflict between Diane and Marcus before a call center complaint spike forced a resolution. What escalation procedure could have brought their disagreement to the steering committee, or an equivalent body, before the flawed homepage ever reached a pilot launch?
2. Design a RACI matrix for the citizen portal deliverable as it stood before the steering committee intervened, with Diane and Marcus both claiming accountability. Who should actually hold the single Accountable role for a homepage decision like this one, and what happens to Diane's and Marcus's authority over their own departments' features if neither of them holds it?
3. The steering committee resolved the immediate conflict, but its authority exists only through the deputy secretary's interim authorization rather than a permanent charter. Is that governance fix actually sufficient, or does it just relocate the same fragility, an authority tied to one person's tenure, one level higher in the organization?