Researching one government experience

Oman had already invested heavily in digital government, and individual services worked. Some were genuinely well regarded, the ROP app came up unprompted, interview after interview, as the standard people measured everything else against.

The difficulty was structural, not technical. Services had been built independently, each inside the ministry that owned it, each with its own terminology, navigation, login and support process. Every one made sense from the inside. Collectively they required the user to understand the government's internal organisation before they could complete a task.

People do not organise their lives that way. They arrive with a goal: transfer my employment to a new company. Register a business. Bring my son to live with me in Oman. Get a visa in time for a wedding.

The audiences shared very little. A citizen fluent in Arabic and familiar with government process. A resident who cannot read the language her own official documents are written in. A first-time business owner in Ibri with no network to ask. A tourist applying from Germany. A man with a hearing disability who cannot use a counter service because there is no interpreter in the building.

They differed in language, digital confidence, physical access, urgency and trust. A single portal cannot be designed around all of those differences individually, it has to be designed around what they share. Finding what they shared was the research problem.

Six personas came out of the research, one per audience group, each grounded in patterns that recurred across multiple participants rather than composited for coverage.

We ran structured five-whys sessions on the strongest recurring patterns rather than accepting the surface complaint. The most productive traced the tendency to abandon the online route at the first obstacle. It did not resolve to poor interface design.

It resolved to an expectation, formed over years, that government processes are inherently complex, combined with the absence of any reliable in-product help, which made seeking a human intermediary the rational first move rather than the last.

Several Omani participants stated explicit confidence that government would protect their data, contrasting it favourably with global platforms. The same participants described avoiding official channels in practice.

The research did not produce a list of interface corrections. It produced the evidence base, personas and journey models that the portal's information architecture and service strategy were subsequently built on: an experience organised around goals rather than entities, consistent across services, transparent about process and cost, and inclusive by design from the outset.

The framework deliverables proved as valuable as the findings. A shared discovery structure let a distributed team conduct comparable research across very different audiences. The personas and journey maps gave researchers, designers, product owners and ministry stakeholders one common reference point, and discussion moved from opinion about features to evidence about people.

The outcome was not a redesigned portal. It was the foundation that made a credible one possible, and a decision-making framework the programme could keep using after the research phase closed.

At national scale, research is not a phase that produces a document. It is the shared language a large, multidisciplinary programme uses to make decisions. Insights that other disciplines could not explain, defend and apply had no effect, however well evidenced they were. Learning to synthesise for use rather than for completeness was the most valuable skill the project taught me.

It also reset where I look for problems. The services we studied were not badly built. They failed people in the space between them, in the moments where nobody had responsibility for the whole path.