Understanding Value Streams | Why Didn't You Test That?
Why Didn't You Test That? The Curiosity Software Podcast, hosted by Huw Price and Rich Jordan! In this episode, our hosts are joined by Chris Dutta, Co-Founder and Head of Consulting at Dragonfly, together they explore the exponential business and quality benefits of implementing value stream management.
Bring Quality to Your Software Delivery
In this episode, our hosts are joined by Chris Dutta, Co-Founder and Head of Consulting at Dragonfly, together they explore the exponential business and quality benefits of implementing value stream management (VSM).
Chris Dutta succinctly describes VSM as a set of activities improving the end-to-end flow, quality, and the value that IT provides to broader business objectives, typically through software delivery!
"Going up the test pyramid, modelling out the building blocks of how a system works is the technical product implementation of a value steam. Then optimising those models gives an independent, objective articulation of complexity. Scoring and story points can be based off that."
- Rich Jordan, Enterprise Solutions Architect, Curiosity Software
"Isn’t this the point of AI, to gradually introduce more metrics to be able to predict? But none of that makes any sense without a model of requirements. Get that working, then you’ve something stable to feed into the tickets, into test cases, and from there measurable impact in the value stream."
- Huw Price, Managing Director, Curiosity Software
"If you get quality right, that improves the whole experience around value streams. We underpin our whole approach to engineering quality in software through Value Stream Management. Everything we do has got to align to business value, business output, quality and flow."
- Chris Dutta, Co-Founder and Head of Consulting at Dragonfly
-
Shownotes
00:00 - Introduction: beyond testing towards better software.
02:17 - Chris Dutta outlines Dragonfly’s mission statement.
05:25 - Chris Dutta distils Value Stream Management (VSM).
07:51 - Why VSM isn't done well at Enterprise level.
09:42 - Misinterpretations of implementation of agile.
13:00 - Going beyond Dora metrics as guiding KPIs.
15:23 - Does business actually know what it wants?
18:06 - Complexity metrics and testability.
21:20 - A model tells complexity at an objective level.
24:27 - Multiple data sets for accuracy and confidence.
26:31 - The role of AI: iterating requirements.
29:28 - Key insight: avoid getting lost in the data.
30:30 - Dora metrics exponential.
32:25 - Neuro normalisation engine.
35:30 - Adopted scaled agile but having difficulty.
38:00 - How is VSM approached?
-
Full Episode Description
In this episode, our hosts are joined by Chris Dutta, Co-Founder and Head of Consulting at Dragonfly, together they explore the exponential business and quality benefits of implementing value stream management (VSM).
Chris Dutta succinctly describes VSM as a set of activities improving the end-to-end flow, quality, and the value that IT provides to broader business objectives, typically through software delivery!
Dragonfly’s product Neuro extracts and understands such activities, surfacing them as metrics into a dashboard for VSM discovery. Through surfacing insight from commits and repos relative to active or blocked ticket states, this impacts on lead time from development to deployment in multiple Jira instances. The outcome is reduced waste in the SDLC. But how insightful certain tickets are depends on a secondary layer of complexity metrics.
While the purpose of VSM is to bring business value to the end user, the benefit of a value stream is not immediately evident. The main question is then how to leverage the metrics without gaming them? It’s known that measurements are insightful, but only if used to inform what the business wants. At worst, a siloed value stream pitches measures as weapons which can deflect from continual improvement.
Inevitably, forces existing in large organisations naturally lend themselves to waterfall-esque scenarios. So, in practice there are many contradictions for product teams to navigate, either from top-down C-Suite colleagues or bottom up from a practitioner perspective.
Ultimately, how do you compose an agile certified product team, that distributes the VSM knowledge and skill sets required across the whole business?