It was close to a goosebump moment. On Wednesday morning, colleagues from the City of Wiesbaden, DigitalAgentur Brandenburg, Digital Service, and PD shared their study data on the practical use of Germany’s digital identity systems. For months, they and others have collaborated on a paper that brings together existing qualitative and quantitative data on digital identity implementation.
Their findings aren’t too cheerful. But they aren’t novel either. There is a 70 to 90 per cent dropout rate among users of digital government services when they reach the common digital identity component. It feels like an open secret, but it isn’t one. Digital teams have been discussing this for years. I see the occasional LinkedIn post about it, too. But no one more senior seems to be discussing it.
So maybe the summary jointly put together by the cross-government working group can shift the debate a bit. Data from 7 previously unpublished studies were included in the paper. Colleagues from 7 public organisations worked on it, at the city, state, and federal levels. It’s a kind of user research cooperation we haven’t seen before.
The findings of the report summarise the shortcomings of the current implementation of digital identity in government services:
- Limited prior knowledge and a lack of guidance cause uncertainty: Users are taken by surprise by the requirement to identify themselves during the process. Texts that are too long and confusing administrative jargon are overwhelming.
- Systems disruptions in the user journey cause confusion and overwhelm users: The process requires too many switches between web pages, devices and apps. Often, after logging in, the automatic redirection to the actual service fails.
- Lacking transparency in security requirements undermines trust in the application: The varying assurance levels applied across administrative services are not clear to users. Missing or lost ID card PINs require a visit to a government office.
- Technical faults and a lack of fault tolerance lead users to abandon the process: The system is prone to errors. NFC scanning on smartphones frequently fails, and various browsers display error messages without clear instructions for resolving them.
- Information overload and a lack of support cause users to abandon the process: There are too many contradictory instructions and videos, but no genuine first-level support when problems arise. Neither the 115 citizen helpline nor the individual government departments can help with app errors.
- A lack of everyday experience with government-provided identity services leads users to prefer private-sector alternatives: Because government-issued identities are not part of everyday life, users often opt for the familiar, private-sector solution when given a choice.
- Technical and formal access requirements lead to structural exclusion: The system systematically excludes people. Anyone without an NFC-enabled smartphone, with outdated software, or – as a third-country national – without a valid EU identity document, is completely cut off from essential digital government services.
Those are a few areas to work on. It’s nothing one entity can address alone. Just as the insights into the problems, the fixes need to involve collaboration and further research equally. It needs a clear target for a significantly lower dropout rate, along with considerable flexibility and experimentation to find ways to achieve it. For that, it demands senior attention and sponsorship.
The recommendations from the report are described along 4 layers, borrowed from an international PwC study on digital identity.
- Regulation and governance: Establish centralised governance, introducing user-centred legislation and removing legal barriers, and evaluate safety standards
- User-centricity: Use binding standards, ensure a consistent brand identity, provide intuitive user guidance, review information provision, incorporate inclusivity from the outset, and establish mandatory end-to-end tracking
- Building trust through standards: Harmonise the user experience of platforms and mailboxes, provide open-source solutions, make authorisation certificates free of charge, standardise for local authorities, allow established alternatives, and ensure mandatory cooperation between providers
- Infrastructure maturity: Failure tolerance and support strategy
Some 90 people signed up for the Wednesday presentation session; over 50 showed up. The paper is published on the NExT website.
I am curious to see if and how the findings will be discussed. For me, the voluntary collaboration on a working level across so many public sector organisations alone is a huge achievement. We haven’t seen anything like this self-initiated user research collaboration in Germany before. It’s evidence for capability and capacity that didn’t exist just a few years ago. It is also evidence of a shared pain with common components that don’t work as well as they should. The people working together on the paper used their energy from a moment of frustration and channelled it into hope that, together, they could change things.
Progressing with training, trust study, and spend controls
Our little digital transformation instrument ecosystem continues to grow and evolve. It’s the constant progress that makes the difference, not the big-bang releases.
This week, I participated in a session on the visual stimulus for the trust-by-design study that the Competence Centre for Public IT (ÖFIT) is planning to run later this year. Designer Sophia made various drafts we discussed with the researchers. She will continue working on these in the coming days while the survey design is developed. We hope to extract reliable evidence that a unified design and brand system, along with service quality criteria, can help increase trust among people in the state.
On another day this week, Linda, Sabrina and I met to catch up on the reworking of the ‘Introduction to service design’ training we are preparing for late summer. We offered the training some 3 or 4 times, once in German. Now, we are weaving the Service Standard points into the training offering, and it is coming together nicely. All but 3 of the 13 points of the standard will be referenced in the training as the various exercises put them into action for the participants. Currently, we are aiming to run the first new iteration of the training in the first week of September.
Also this week, I checked in with colleagues working on IT spend controls for the German Federal government. As I wrote in January: Standards won’t do anything unless integrated into a larger system that includes carrots and sticks. That is why their work matters so much. I shared some of the things I have seen work well abroad and offered my support where desired.
What’s next
It will be a somewhat shorter week, in some ways. I will have a Focus Friday on the train and am eager to make good progress on blog posts.
There is the blog post on open roadmaps that I started preparing with product manager Bene in the spring, which I want us to publish this month.
And there is a lunch-and-learn session on our user-centred design discipline on Thursday, which we plan to plunder for a blog post, too. Charlotte, Daphne, Sonja, Torsten and I will be talking about our sub-disciplines, what connects them and sets them apart. We will have content ready for Thursday, so I might be able to turn some of that into a blog post on Friday. I expect that to be an easy write.