How To Improve Quality

Beren Van Daele, The Core of Things

“Give me six hours to chop down a tree and I will spend the first four sharpening the axe.”

― Abraham Lincoln

What’s next?

Another famous quote is: plans are worthless, but planning is everything. I’d argue that following through is at least as important. We used to only have 3 phases of RiskStorming. The Online version introduced a 4th: Commitments, Deliverables & Measurements.

We try to achieve two things in this phase:

  1. Broaden the backlog past Functionality
  2. Plan little steps to move forward on our Quality Journey
User Stories typically deal with functionality.

Past Functionality

Back in Phase 1, you identified 6 Quality Aspects of which Functionality is or isn’t one. In my experience, the team’s backlog is filled with functionality stories. There might be an odd security, performance or accessibility story or acceptance criteria or requirement interwoven in it, but that’s rather rare.

By picking at least 5 other Quality Aspects, identifying their priority issues and discussing possible mitigation steps you virtually built up several new ‘backlogs’ of work items that need some attention. Now I’m not suggesting you create 5 separate backlogs, or worse yet, start silo’ing these things up further in teams or roles or,… Please no.

Pick the highest priority issues, something of a minimum viable product and integrate them in your backlog/kanban board/requirement list,…
Don’t try to complete everything, you can’t, but tackle the most pressing things first. Convert these requirements/plans into work items for your backlog.

Move Forward On Your Journey

At the very beginning of your RiskStormingOnline workshop, you discussed where you are on your journey towards the a goal. In this stage, you can decide together how you’ll achieve the next step. Are you worried about performance being an issue?

  • Analysis: You might first want to define benchmarks for performance. What’s acceptable? What isn’t?
  • Operations: Can we check what the current average times are? What are bottlenecks?
  • Testing: How can we measure safely, smartly and reliably how the system will react under more stress, persistent load and large number of requests?
  • Development: Can we optimise our systems to make them more fluent? Do we need separate testing environments? Would they give realistic results?

Start with defining benchmarks and measuring what the current results are. Maybe ask a tester to explore and investigate potential problems and report back. Once you get a feel of whether things are satisfactory or not you can implement measures and drill deeper into the new problems, if they matter enough.

See how this ties into the heuristics, planning, strategising discussion from phase 3? Just like the squad that was caught behind enemy lines: First get your bearings, evaluate and scan your possibilities before taking a well thought-out action. …or take action and be ready to deal with the consequences.

How To Act For Quality?

Phase 4 is here to make things concrete. The discussions we have gone through to get here were invaluable. They gave you purpose, vision and means to achieve your goal. Here’s the next step to make them actionable, trackable and measurable.

  1. Make commitments
  2. Provide means to stay accountable
  3. Visualise your journey

Make Commitments

In Phase 4, we structure the information from the previous phases into ‘streams’. This means we see combinations of one Quality Aspect, a risk and it’s possible solutions. This gives you a great overview of what your team discussed.
From here, people or roles can make commitments to take action.

Compare it too action points following a retrospective. A retrospective without action points is a complaining session. A RiskStorming Session without action points can give direction, but has no plan.
Phase 4 helps you translate your discussion into actionable tasks, user stories, sticky-notes,… Any body of work you normally work with.

What you’ll be looking for is commitments from teammembers to take up part of the work that was discussed. Raising possible solutions but waiting for others to achieve them doesn’t help at all.

Accountability

As far as possible, define what success would look like. How does the person making the commitment and other teammembers know when they succeeded in their commitment? What are they measured against?
Measurements have a bad reputation in IT, but they can be very helpful when applied specifically and when the team carries them. In other words, when the team defines the measurements together, the measurements are valuable and acted upon.

Measurements can be numbers, but also feelings, happiness, an achievement, milestones, KPI’s, OKR’s, the result of a questionaire,…

Goodhart's Law for Data Science and what happens when a measure becomes a  target?

Visualise Your Journey

Over the years, I’ve experimented with several ways to visualise where we are with our quality.
I encourage you to be creative and try your own ideas, things that make sense in your context.

The important factor here is that you find a way to make your progress visible to people within the team. This might be more difficult in a remote setting, but with various communication tools you’ll find plenty of ideas to share your charts, models, walls,… I’m sure.

Here are a couple of ideas:

A one-page-test-plan to capture the most important ideas on one page. Especially highlighting issues the team may have which they can’t solve themselves.

A RiskMap to hang on a wall to show phase 1 and 2 of RiskStorming. Buy some red and green stickers to highlight on the map what is going well and what isn’t.

A wall of bugs with actual pictures of bugs to indicate the severity. Butterflies are trivial, scorpions are blockers. Start exterminating!

What’s next?

Hopefully, this has been a wonderful exercise for you and the team. Not only did you learn about how to deal with quality and risks, but you applied it to your current project and your current situation.

Among many other benefits, these are probably the

  • Work that was previously invisible are considered wasteful is now brougt out in the open.
  • You can track quality in a more meaningful way
  • Whenever a new teammember joins, walking them through the results will be very enlightening for them.
  • The shared understanding of your product and project by your team has greatly improved.

Now is the time to make the next step on your journey. I hope you feel much better prepared now.

Virtual Tolkien Is A Free VR Recreation Of An Iconic LOTR Film Scene
Good luck on your journey, traveller.

One thought on “How To Improve Quality

  1. Goodhart’s Law: reminds me of a Soviet-era cartoon, depicting a mammoth nut and bolt being pulled out of a factory by (in comparison) a tiny, tiny tractor, whilst the manager announces “Thus, in one step, we exceed the targets of the Five Year Plan!”

    Like

Leave a comment