Methodology

HackIT Agile Lifestyle

We followed the HackIT Agile Lifecycle developed by Hackney Council in order to build a digital service so good, people prefer to use it.

Multi-Disciplinary Team

• User Research - organising and running user interviews, shadowing and testing and relaying results back to the team.
• Service Design - interpreting the user research and translating into user stories and wireframes
• Delivery - establishing Agile procedures, coordinating work prioritisation and managing quality
• Technology Lead - responsible for the technical framework, standards, testing, security and back-end development of the service
• Front-End Developer - coding front-end to wireframe specs and working closely with back-end development to integrate APIs and data
• Service Manager - leading on day-to-day operations of the NCC and coordinating testing and roll-out with agents
• Product Owner - advising on requirements and ensuring MVP meets the needs of the NCC
• Service Owner - maintaining the overall vision of the service

Capacity

We were able to manage resourcing across the service fairly well. In Discovery and Prototype, there was more of an emphasis on user research and service design resources and three people were required but we were able to reduce this two in Build stage. This was in contrast to Development, which tends to weight towards Build, for obvious reasons. However, the Technology Lead was required throughout the phases as we needed to understand technical constraints and requirements and make CRM decisions early on. The NCC stakeholders were available for advice and decisions, however, as the NCC is essentially a live event happening day in, day out we did have to make concessions to operational norms. This particularly applied when we needed to test with actual agents and we had to closely coordinate schedules so that we did not poach staff during busy periods.

Software

• Trello for tracking all other tasks outside build - we maintained a Kanban-style structure with To Do, In Progress, Awaiting Response and Done.
• Axure for hosted, clickable wireframes that were suitable for hi-fi testing and proof-ready for coding.
• Jira for managing development of user stories, backlog management and sprint planning. We used one Board to track Design and a second for tracking Build. Tickets between each were linked so that the front-end developer knew which wires were approved and Ready for coding.
• Slack for day to day team discussion and quick answers to queries.
• Google Team Drive for shared asset storage.
• This website for onboarding new team members and showcasing internally and externally.

Schedule

Discovery kicked off at the beginning of May and proceeded for around eight weeks. This was followed by around four weeks of prototyping, which resulted in a set of hi-fi wireframes. Sprints were brought into the project for the Build phase, which began at the start of August. Sprints were generally two weeks and there are eight sprints in total. The site was deployed to live first on 25 October 2018 to a small group of agents, with limited functionality. The full MVP is intended to launch in the third week of November.

Agile Ceremonies

Across the schedule, the team has held daily stand-ups to identify blockers and establish progress within the different disciplines. These were attended by the multi-disciplinary team of delivery, service design, user research and development throughout.

We have run three Show and Tells: in the middle and end of Discovery and again at the end of Prototype and start of Build. In between, the team at the NCC was updated on a regular basis via Week Notes as well as informally whenever we sought clarification or advice and during any testing we carried out with NCC agents. We also ran additional demos to them when we needed specific feedback. A wider audience can view our Week Notes on HackIT Delivers G+ as well as on Pipeline and on this website you're looking at.

We held two retrospectives: one at the end of Discovery and a second after Build sprint 4.

Before each sprint, we organised backlog grooming and sprint planning sessions. We set a goal for each sprint and prioritised stories that would meet this goal. For example, Sprint 7 focused on taking a rent payment and viewing transactions.