Public sector digital transformation programmes tend to start in the same place: a new portal, a cleaner form, a promise that citizens will no longer need to phone a call centre or visit an office to get a straightforward service done. This is good ambition and it is also, on its own, an incomplete plan, because a front door only works as well as the building behind it.
The building behind it, in most public services, is a mix of legacy case management systems, manual verification steps, and staff who have built years of informal workaround knowledge that no digital form captures. A front door that submits cleanly into that back office does not speed up the service. It just moves the queue from a phone line to an inbox.
The front door is only as honest as the back office
A digital service that promises a decision in five days but sits on top of a back office process that genuinely takes three weeks is not a faster service. It is a service that has moved the delay somewhere less visible, and citizens notice the gap between promise and reality faster than programme teams expect.
Good digital front door design starts by mapping the back office truth before designing the front end: what actually happens to a submitted form, how long each step genuinely takes, and where the manual bottlenecks sit. That mapping is unglamorous and frequently uncomfortable, because it usually reveals that the back office needs redesigning at least as much as the front end does.
A front door only works as well as the building behind it. Citizens notice the difference faster than programme teams expect.
Assisted digital is not an afterthought
A meaningful share of citizens using any public service will not, or cannot, complete a digital journey unassisted: through lack of access, lack of confidence, complexity of circumstance, or simple preference. Treating assisted digital support as a fallback bolted on after the main channel is built consistently produces a worse and more expensive service for exactly the citizens who need the most support.
The services that get this right design the assisted route alongside the digital one, using the same underlying process and the same data, so a citizen helped by phone or in person gets the same outcome and the same speed as one who self-serves online. That parity is a design decision, not a resourcing afterthought, and it needs to be made at the same time as the portal is specified.
- 路Map the true back office process before designing the front end
- 路Set the digital promise to match the slowest real step, not the fastest possible one
- 路Design assisted digital as a parallel channel, not a fallback
- 路Measure the service by outcome for citizens, not by portal traffic
Measuring the wrong thing
Portal adoption rates and digital transaction volumes are easy to report and easy to present as evidence of success, and they measure channel shift rather than service quality. A citizen who is pushed onto a digital channel and still waits three weeks for an outcome has not received a better service, whatever the adoption dashboard says.
The more honest measure is end-to-end time to resolution and citizen effort, tracked across every channel a service actually uses, not just the new one. Programmes willing to report that measure, even when it is unflattering in the first year, are the ones that end up improving the service rather than just its shopfront.
None of this is an argument against digital front doors, which remain a genuine improvement when they are built on an honest back office and paired with real assisted digital support. It is an argument against treating the portal as the transformation, when the portal is usually the smallest and least risky part of the work.
The public services that earn lasting trust are not the ones with the most polished form. They are the ones where the promise made at the front door is actually kept by the process behind it, for every citizen, regardless of which door they used to get there.