Unboxed Event notes: "Unstick your digital products"


I’m at a speaker session from Unboxed consulting

Dave Evans - Product Manager (E-commerce) at Macmillan Publishing Group

“Trust me, I’m a product manager”: case study from publishing

  • Sales thinking: identify influencers, budget holders etc.
  • Commercial products: it’s all about profit-and-loss. Non-commercial products need different KPIs
  • Product manager != project manager. Responsible for:
    • Backlog, vision statement
    • Roadmap: what questions do we need to answer in the next windows of time
  • “You need to go rogue” - prove concepts on your own without a team
  • 404 test = e.g. buy now button - show it to a % of users, does nothing
  • Concierge test = build front-end, but handle the back-end manually
  • Everything in the backlog should be in line with the vision statement
  • Question for unpicking feature requests which you think might not actually be about solving a problem: “What happens if you can’t build it?”

Glyn Parry – SH:24 (NHS collaboration)

How designing a new experience in the NHS helped to unblock and unleash new potential

  • His experience: People in localgov/NHS aren’t familiar with agile
  • Just because they don’t think of thinks in agile terminology doesn’t mean they don’t have an agile mindset
  • Sexual health services in Lambeth&Southwark are very overstretched, and in rural areas, round-trip times to visit a clinic might be very long
  • Online STI kits - convenient but expensive
  • New service: free, user-focussed
  • User journeys illustrated with comic-strip style graphics: @g_parry_24 talking about SH:24 approach
  • Identify people in the organisation who have an agile mindset and are quite influential
  • Approach: agile prototyping, building a service piece-by-piece with users
  • Developing personas
  • It’s a digital project, but only 25% of the service being provided is digital - needs to be recognised
  • Started with the basic GDS form
  • Prototyping the kits that people recieve - cardboard and post-its
    • Started out trying to design engaging packaging for the test kits
    • through user testing: learned that people don’t want this - they like a really simple approach
  • Need to think about the user needs of the people receiving the tests as well
  • Continuous evaluation on multiple different angles: measuring the impact of the service on sexual health
  • Assumptions broken: people don’t actually want their results to be super-discreet: they just want to know as quickly and clearly as possible
  • Prior to alpha - 3 month discovery stage: very intensive working with users

Will Rowan - Product Manager at the Ministry of Justice (interview)

  • Managers are more important if they have more responsibilities - therefore they’re predjudiced towards bigger projects and bigger products
  • PRINCE2 documentation is large regardless of project size - again predjudicing larger projects
  • Project example: Multi-car insurance. users not considered, data produced not considered, organisational capability to run the service, not considered (!!)
  • On the importance of competitors: “If you’re trying to motivate a project team you need an enemy”
  • “The more that you can get product in front of people the better” - internal and external. The importance of show and tell. “we can show you what’s possible” - unlock stakeholder committment: low cost, low-risk “If they can quantify the risk they’ll sign it off”
  • Stakeholders probably don’t believe that you can deliver that much in that little time. Chip away at this by sharing what you’re doing and how you’re doing it.
  • Drew the whole of a product on one A2 flipchart - helped to convince stakeholders it’s feasible
  • Attraction of Waterfall is that it lets you sign off the whole budget. In agile you’re iterating the budget as well as iterating the project
  • Getting everyone speaking the same language is important - e.g. training stakeholders to not refer to things as ‘portals’ when they’re nothing of the sort
    • If you discover that you’ve got it wrong, or there’s a better term, it is worth changing it all the way down the stack
  • When running agile: Make the work visible - extremely important: whole backlog on the wall. Makes it much easier to have a common understanding. Always carry whiteboard markers and post-its when in a non-agile environment (you won’t find them there).

Richard Stobart – CEO, Unboxed Consulting, Agile Coach of the Year 2014

Techniques for overcoming the Big 7 digital product blockers

  • The trap of 100% utilisation: Can’t move quickly. Grind to a halt.
  • 70% utilisation is about as fast as you can go: helps avoid context-switching
  • Intuitive thing when you see someone not busy is to make them busy, but this slows down all their other projects
  • If you’ve got spare time, help unblock other people, or do nothing”: don’t add more things to the pipeline
  • Always be delivering the highest value thing on your most important project
  • Optimise ‘batch size’ - grouping together features for deploy. Small batches = responding to change more quickly. = fixing bugs more easily: if the problem is a result of code which was written recently, the devs have that conceptual stack in their head.
    • Not achievable if your deployment process is onerous or slow
  • People used to think of environments like pets: they all had names. Today they’re like cattle: set up when you need it, tear down when you’re done.
  • rightscale + ansible + docker: used to spin up 1000 test servers to test releasing a product to all students in China (which was the MVP!!)
    • Cost £10k to run that test but was trivial to set up
  • Creating roll-back-able database migrations: dbdeploy, liquibase etc. - have both database schema versions live in parallel, delete the old one when the new one is stable
  • In closed-scope projects, everyone feels like they need to throw in all their great feature ideas before the scope window closes - leads to feature bloat. Solution: get stakeholders on side, and ensure that the product owner controls the backlog