28 Nov
Posted by: Michele Ide-Smith in: LocalGov, User experience
I work in a team that manages local government web development projects. We work closely with applications developers in IT and we do most of our development in-house. A couple of years ago we realised noticed a pattern in the way we were working. Some projects had complex requirements and often involved working with new technologies. Bigger projects took 12-18 months to complete, which meant that work on the other web sites or applications had to be put on hold.
Early in the project the teams would have workshops and lots of meetings with colleagues in the relevant council services (i.e. business areas). The project managers would develop very detailed specifications to document the requirements. We would produce wireframes to show how we proposed the interface would look (and sometimes we tested the wireframes with the site users). We would send these lovingly crafted specifications to our development team who would spend a few days reading through our bloated documents, trying to interpret what we meant. The development team would then work up a detailed estimate for the work. Between us we would spend a considerable amount of time negotiating scope and estimates before agreeing to proceed with development.
At last the developers could roll up their sleeves and start writing some code. The developers would work tirelessly to develop functionality that met all the requirements in the specification. Sometimes the developers weren’t able to complete all the functionality because they encountered problems they weren’t expecting because the technologies were new to us all. When the services got to do acceptance testing, they would ask for additional or different functionality, but by this time we’d used up the budget and run out of time to do any further development. If you hadn’t already noticed, I’m talking about the waterfall development methodology which follows a sequential process as follows:

We needed to find a way we could work more efficiently, improve communication with the development team and services and meet business and user needs more quickly.
We’re about to change the way we work quite dramatically. We are proposing to use an Agile methodology known as Scrum. The Scrum Alliance defines Scrum as:
A team-based framework to develop complex systems and products.
This means working collaboratively in a small team on short, “time-boxed” iterations of design and development with daily face-to-face communication in the form of a Scrum meeting. The aim is to turn prioritised requirements (or “user stories”) into working software, but to be adaptive to changing requirements throughout the development process. The Scrum process looks more like this:

Agile is sometimes referred to as a ‘philosophy’, or ‘state of mind’. But local government is not really known for being agile. I often hear people refer to how long it takes to get simple things done in local government.
So how can we best fit the principles of working in a more Agile way within local government constraints? Well we’ve already overcome some of the constraints in making a decision to try using an Agile development methodology, instead of the more traditional Waterfall methodology. I’m going to focus on two areas that we are likely to present us with the biggest issues as we start using Agile for real on our web development projects.
One of the principles of Agile states that:
The most efficient and effective method of conveying information to and within a development team is face-to-face conversation.
Another is:
Business people and developers must work together daily throughout the project.
Agile requires frequent face-to-face communication and team collaboration. Some of the barriers that we might face include:
The second area I think will pose the biggest challenge to us will be designing the user experience in an Agile way. By ‘users’ I mean citizens and have deliberately not referred to them as customers. In Scrum the ‘customer’ is usually the stakeholder who represents the business (and often controls the budget). I have written a couple of previous posts on how we are adopting user-centred design (UCD) methods within the Council.
In the past we have produced wireframes to communicate design ideas and, where possible, tested them with users. But wireframes are time consuming to produce and are often produced in isolation, so only reflect one person’s ideas about how something should work. This has caused us problems because we’ve treated the wireframes as ‘blueprints’ and get services to sign them off. But, as Brandon Schauer points out, a wireframe is a static design of a screen and doesn’t convey interactions well. So as a result what we’ve designed in a wireframe can’t be implemented by developers straight off and developers often have to make design decisions in isolation. Ideally we need more two-way communication and collaboration. So we’re going to have to find quicker and easier way to create designs and work as a team.
Techniques like sketching and paper prototyping could help us to rapidly generate design ideas (see Bill Buxton’s book on sketching and Carolyn Snyder’s book or the new Todd Warfel book on prototyping). I’m keen that we can try using sketching and rapid low-fi prototyping as a team, but I think we have a few barriers to overcome. For example:
I don’t think any of these barriers are insurmountable, but we’ll be on a steep learning curve. Hopefully I can share our experiences in a few months when we’re a few sprints wiser. I’m curious as to which barriers will prove to be the most significant and what solutions we come up with to get around them.
This content is published under the Attribution-Noncommercial-Share Alike 3.0 Unported license.
11 Responses
David Wilcox
28|Nov|2009 1Thanks Michele for such a useful peak behind the curtains at the challenges of co-design processes. I think that the challenges of the philosophy – and the practical obstacles – are played out in many non-tech settings as well. So – do you pick off the problems one by one, or try and get a culture-shift that makes this approach easier? Or both?
Harry Harrold
28|Nov|2009 2We’re about to start our first cross-departmental localgov project – and expect take our Agile working methods with us.
Thanks for the heads up on a few challenges… We’re geographically near them, so maybe meetings should be at our post-it-festooned geekhaus but I’ll be suprised if we manage daily face to face contact. I have visions of exposing our sprint backlogs and encouraging the localgov folk to reorder tasks we’ve not yet started to try and help with this;
I’m worried that our paper prototypes will be seen as simply a more interactive wireframe – as a blueprint of what will be, rather than our best concept of what will be at this point in time…
Thanks for the insight – very useful and very timely.
Michele Ide-Smith
28|Nov|2009 3David – in response to your comments, both I think! We’ve got to solve the practical barriers somehow to remove immediate obstacles, and we have some ideas about how to get round them. But adjusting the culture to accept ‘co-design’ will take longer. I’m hoping we can use some design games like design consequences to get into the swing of it!
Michele Ide-Smith
28|Nov|2009 4Harry – I would suggest you are very clear with the client that you need someone from the business to input regularly to the sprints. And make sure you explain the purpose of prototypes clearly. Our services expect us to get on with the development once we’ve had initial requirements workshops/meetings. I’ve always liked this cartoon which explains how business requirements get mis-interpreted along the way. We’ve also struggled to convince services to part with budgets up front when we have no concrete specification to show them. Good luck with your project!
Harry Harrold
28|Nov|2009 5Yes, will do – business were deeply involved in the purchasing decision and were at the show-and-tell for sprint methodology and paper prototyping we did during the Xth part of the purchasing process so we’re at least reasonably confident there at the moment. At the moment… As you say – regular/frequent contact is key.
Love that cartoon too – never seen that particular version.
PaulGeraghty
28|Nov|2009 6Came across this last week;
http://www.uistencils.com/browser-sketch-pad.html
Which might be helpful to you.
There is one other force at work which you do not mention, probably because its so bleeding obvious, but in Local Gov I had to point it out like ALL THE TIME – that month by month, incrementally and oh so gradually, the web changes like shifting sand beneath our feet.
What was considered cool a year ago can be passé today, the predictions the pundits were making 2 years ago have been overtaken by technology, networks or browser makers.
What your DMUs (Decision Makers) read in the last SOCITM Better Connected report were old hat well before they they were written…
Michele Ide-Smith
29|Nov|2009 7Cool thanks for that. I like the pixel ruler! I came across the Adaptive Path templates last week but in the past I’ve just taken a screenshot of a browser, removed the middle bit in Photoshop then printed out copies to use as a template for sketches and prototypes.
You’re right about the constant changes in the web world. But that’s what I like about it. Lots of challenges. The reality in local government is that you’ll always be running to catch up. Local gov web sites are certainly complex beasts, not least because the users and their goals are so diverse and you may have a whole host of disparate applications to knit together into a supposedly seamless user experience, although many of them will not be web standards compliant.
Steph Gray
29|Nov|2009 8Great post, Michelle. I’m stuck in rather a waterfall project at the moment, which would be much better managed using Agile principles, I suspect.
The challenges for us I suppose include procurement (the detailed requirements are often specced before the developers are appointed); distance (our developers are hundreds of miles from us); and IT culture (I’m guessing, perhaps wrongly, that our project manager from corporate IT would look askance at an Agile approach)… would be interested to hear your experiences as the project develops.
Michele Ide-Smith
29|Nov|2009 9Hi Steph, Agile may not be the answer to all your projects. If your organisation allows (and has facilities) you could use video conferencing or remote software to review what has been delivered in sprints. But for co-design you’d really need to be co-located with developers. The whole team need a shared understanding of the Agile methodology you choose (e.g. Scrum, DSDM). We got an external consultant to come and do training, but it’ll be up to us to sell it to Council services.
Re. procurement – in theory you would define and prioritise requirements up front (a product backlog in Scrum) then get a fixed cost for a number of sprints that would deliver the highest priority requirements. That gives you flexibility to change and re-prioritise requirements between one sprint and the next without incurring additional costs. But be prepared not to get all your requirements delivered in the sprints.
We’re Scrum newbies so we’ll have to see how it goes
Bookmarks for November 27th through November 29th
29|Nov|2009 10[...] Barriers to Agile web design and development in local government by Michele Ide-Smi… – Great post from Michele about the challenges of designing and developing within large organisations – and the opportunities presented by agile methods. [...]
links for 2009-11-30 « Working Notes 2.0
30|Nov|2009 11[...] Barriers to Agile web design and development in local government by Michele Ide-Smith excellent resume – not just of the barriers but of Michele's intended agile approach (tags: egovissues techniques technology) [...]
Leave a reply
About me
Archived posts
Find me online
Gov 2.0
User Experience
Various friends
Subscribe:
Tags
My bookmarks
My Comments
Great notes Sharon! You’ve captured the main points really well. I was obviously doing too much talking and not enough note writing…
My latest tweets
A design creation of Design Disease
Copyright © 2007 - Michele Ide-Smith - is proudly powered by WordPress
InSense 1.0 Theme by Design Disease