Redevelopment of newzealand.govt.nz

Hello. I’m Jared Gulian. I’m the product owner for the redevelopment of newzealand.govt.nz.

[Image projected on screen: Screenshot of newzealand.govt.nz homepage.]

This is the site in its current form. It was last worked on in 2008. It hasn’t had a lot of iterative development since then. And it is based on user-centered topics. But we know from user engagement analysis that only a third of the users are actually getting value out of the site.

So we’ve got a problem that we’d like to try and solve. And it’s a domain-wide problem. It’s not just specific to this one site, but it’s about the system of how government delivers information and services online. We know there’s a lack of collaboration between agencies. There’s some exceptions to that. But generally, agencies deliver information saying, “I’m department A. I do these things.” and “I’m department B. I do these things.”

The problem is that the user journey often requires people to interact with multiple agencies. So getting that high level view is the challenge.

[The presentation temporarily stops being projected on the wall.]

OK, I’ll be very careful not touch the podium.

So we know that there’s a lack of collaboration across agencies. And they’re not delivering information — DIA is also responsible for this — not delivering information in a user-centred way. Information and services are often delivered according to the structure of government, not the user needs. And sometimes because of that lack of collaboration and siloed thinking, there’s duplication across information from different agencies.

And often we know from our user testing that people are having a difficult time finding the information that they’re looking for online.

So we’ve got some basic project principles for how we want to approach the redevelopment of this site. We want to base our decisions on the needs of the users, not the government structures. We know that we need to have an evidence-based approach. So we’re user testing. We’re doing research. We’re getting feedback.

We need to start small and iterate. We won’t develop a huge thing and deliver it on day one, but we’ll start small and we’ll go from there.

We know we need to create a consistent user experience. That’s one of the key problems that people have when interacting with the government online system. Because every website is different, and it’s often difficult to manage their way around the space. We know that we have to collaborate across agencies to deliver something that’s user centred. DIA cannot do this alone.

We’ve taken a look at what’s happened overseas. You’ll see two websites on the screen. There’s sa.gov.au. That’s the South Australian website. Both of these websites are delivering information in user-centred terms, but they’ve got very different models behind the scenes.

Very quickly, South Australia is delivering what’s called a franchise model. They are actually having agencies collaborate on content and then deliver that in a user-centred way. It goes through a core team and then out into the website.

On the other side, you’ll see gov.uk. They are also delivering in user-centred terms, but they’re doing something that’s much more of a command and control, where they have something like 200 people working on this website, creating content that’s then checked by external agencies. And then they’re delivering it centrally. So these are two very different models.

The gov.uk’s got a much bigger scope than what we’re dealing with. They’re actually trying to aim for a single domain for government. In that Inside Government section, they’re actually bringing corporate content into the site and closing down websites. So it’s a very different and very significant scope.

At this point, I will turn over to Nathan Wall, who’s our information architect and who deals with our user testing.

[Nathan begins speaking.]

Good afternoon, everyone. Jared’s taken you through an example of what the overseas jurisdictions are starting to show us is ideas for good practice. And we’ve done some research to understand what the current site is like. But actually, that doesn’t give us the answer of what do we actually need to produce that works for New Zealanders.

So we thought we’d do something crazy and actually ask people. We’ve run some focus groups. We’ve done various bits of research online using some online tools to ask people how they think about government information and how they would group it and categorise it.

[Image projected on screen: 18 content categories arranged in a circle around a laptop computer that says govt.nz on the screen. The categories are business; community, arts and recreation; crime and justice; driving and transport; education and training; emergencies and disasters; environment, resources and energy; families; government in NZ; health; history, culture and heritage; internet, media and communication; money, benefits and tax; travel and immigration; work, jobs and employment.]

This image is a representation of that initial categorisation. There are roughly 18 groups of content here. And you’ll see looking on the screen, people are treating government information and grouping it by terms like driving and transport, history and culture, families, travel and immigration. There’s no real grouping by government department here. It’s not how people actually think.

So we’re going to use groupings like this to actually start categorising government information in a slightly different way to what we’ve done in the past. All to help us get that working in a more consistent way, we decided we were going to build an alpha site. Now what do I mean by an alpha site?

Well it’s actually a prototype. It wasn’t really a finished full site. It wasn’t fully designed either. It was basically looking at the content model, how do people move through the information space? How can we actually make sure that people can find information that they’re looking for? And it gave us an opportunity to just learn a little bit more about user needs and how people thought.

We did very little visual design of the site whatsoever. Anyone who has done wire framing or sketching sort of designs before, it was almost a plain sight with very, very little visual design.

Then we actually put in front of a whole bunch of people. And we tested them. We gave them a series of scenarios, real life tasks that people would need to do with government. And we actually observed how they moved through both existing websites that government has today and how they would move through a possible alternative, the alpha site that we had built for them.

And no surprise, in some areas they said the alpha site was actually better. One of the scenarios we had people do was look for information on help paying rent. Again, they all went to Google to start that task. We kind of expected that. The problem that they then had was where did they go next after going to Google to do their search.

There’s roughly 14 billion pages of content on the internet. And if you do a search for “help paying rent,” you get 85 million pages of content coming back. And only one of them in the top 10 search results was actually a government page from a New Zealand department. So they didn’t actually have the right starting point.

The next challenge they had was once they’d found a starting point, because each site was built in a different way, they actually had trouble finding what they were supposed to do. We saw examples of they’d get three quarters of the way through a task and the next step was download the form.

But because the designs of the site were different, they couldn’t actually find the form to download. And it was at that point that most of the users we were interviewing gave us some very strong feedback and a reminder that relying on PDFs, particularly for forms, is still not an online service.

They might be able to find the form. But then a lot of people we were talking to said “I don’t have access to a printer, so what am I supposed to do with this now?”. One of the guys we were talking to said, “Look, I’ll send this via email to my niece. She can print it out for me and post it back. Oh, maybe I need to get her to do two copies in case I make a mistake.”

And another person was saying, “Well, I’ll ring the department. OK, I’m going to spend 40 minutes on hold, and I don’t know if they’re still going to be able to help me.” It wasn’t an experience they were actually looking forward to. So we’ve got to come up with a way of fixing that.

If any of you are interested in some of the research that we have done, we are also taking a slightly unusual step for government — we’re sharing everything. All of our research is being published, and it’s available on the Web Toolkit website. You can see the test plans. You can also see a summary of the actual test results. We worked with a company called Optimal Usability.

Now, we’re actually going to build a beta site. I’m going to pass back to Jared, and he’s going to tell you what we’re doing.

[Jared begins speaking.]

So we’ve finished the alpha site, and now we’re working on the beta site. The beta site will be made publicly available. It’ll be available later this year so that we can actively get feedback from it. It will have thin content which will focus on the user journey and user-centred information. And it will provide that context when there’s multiple agencies you have to deal with in trying to complete your user journey. And it will link out to other agency websites.

We’ll have an iterative, evidence-based approach, as we’ve done with the entire project so far. The existing site will stay live when the beta’s live, and there’ll be a warning saying this is a beta site for authoritative information. Go over to newzealand.govt.nz.

[Image projected on screen: Screenshot of alpha version of the govt.nz website that reuses the look and feel of the gov.uk website.]

This is our content that we developed during the alpha phase, which we’ve put into the gov.uk front end templates. The great thing is the UK have made all of their code open source so we can steal it. So that’s what we’re doing. So we’re taking their front end templates and we’re using them, we’re modifying the templates for our situation. But the fact is, they’ve done a lot of testing on these templates. They have improved them quite a bit. So we can take them and go from there.

[Image projected on screen: Screenshot of the “Driving and transport” page from the alpha version of the govt.nz website.]

This is an example of an information hub. There will be about the 15 to 18 different user-centred information hubs. Again, it’s the UK code that we’re taking. And we’re iterating off of that.

[Image on screen: Screenshot of the “Help with rent and accommodation costs” page from the alpha version of the govt.nz website.]

This is an example of what we’re calling a sign post page. This particular page has information about eight services from six different agencies. Incidentally, the UK site just won an award for its design. So we’re very happy to have their templates.

[Image projected on screen: Screenshots of tablet and mobile phone views of the “Money, benefits and tax” page from the alpha version of the govt.nz website.]

One of the great things that they’ve used responsive design. So what you see there are on the left is the way that site will appear in a tablet view. And on the right is a mobile phone view. So that you don’t have to do a separate mobile site. It responds to the device, and so it will respond to whatever device people want to use.

[Image projected on screen: Screenshot of Department of Internal Affairs directory page from the alpha version of the govt.nz website.]

We’re folding in the government directory into the site. So this is the DIA directory page. There’s currently a public sector directory which is not well known or used. We’re taking that information, and we’ll be revealing it through govt.nz.

This will have information about the services agencies use. It will have information about the ministers and management team and also contact details. We’re working with agencies quite a bit. We’ve received approval from the Digital Services Council in principle to actually do the fact checking that we’ll need before the site goes into production. We’re developing the content in the first instance, and we’re asking other agencies to fact check before it goes live.

We’re working with the Result 10 team and looking at their priority services and making sure that we’re covering those on the redeveloped website. And as I said, we’re asking agencies to fact check that content before the site moves into production.

For users, they’ll have quite an opportunity to give us feedback on the beta site so that we can improve it. They’ll get a more consistent user experience, which will help them to interact with government more easily. It will be easier for them to find the starting points in government. And it’ll be plain English content, so it’s easier for them to understand.

All of our research is available on the Web Toolkit, so that’s webtoolkit.govt.nz. And you can go there and have a look, read more about the project. And feel free — there’s contact details on that website as well — feel free to get in touch with us if you have any other questions.