Value Stream Management (VSM) Implementation Roadmap
This last week was the release of the first State of Value Stream Management Report, which, among other things, provides a how-to guide on implementing digital value streams.
One point the report makes is that organisations often believe they are not mature enough for value stream management. In my experience over the past several years working with organisations to optimise their value streams, I find this to be true, but it need not be a blocker. For example, when discussing data-driven approaches to Governance, a typical response I hear from leaders of large organisations is that their data isn’t in good shape, which they see as a blocker. As a result, they believe they need to execute a cleanup or restructure/re-architect their Agile Portfolio Management tools before they can even begin.
Other common responses include that their teams aren’t good at keeping their systems up to date and accurate and have doubts about their maturity to use analytics platforms that leverage data to produce actionable insights.
The report’s position is that organisations should start measuring their value streams immediately, wherever they are today. That’s sound advice, which aligns well with Kanban principles that encourage organisations to start with what they do now and agree to pursue incremental, evolutionary change.
The report states that even if you are practising waterfall approaches, full of manual activities, you need to benchmark yourself to understand where you are, select improvement targets and have visibility to see if there was a positive change. Waiting for significant transformations or utopian perfect data will only hold you back.
Again, my experience matches this. I’ve had several occasions where my clients said, “let’s talk again in a few months, as we’re restructuring our Jira and will have better data to use soon”. Yet, six months later, they are bogged down in a Jira restructure and still haven’t taken the first step towards becoming more data-driven.
On the other hand, those brave enough to take the first step, regardless of the current state of their data, have experienced a sharp improvement in a relatively short period.
As an example from my experience, a large program running without paying too much attention to flow decided to look at how their delivery performance for new customer features had been over the last years. This is what they discovered:
- Lead time (85% Percentile): 190 days or a bit more than six months
- Predictability (in relation to service level expectation): 42%
- WIP Age (85% Percentile): 336 days or a bit more than eleven months
- Throughput (over last six months): 27 new features
- Overall capacity dedicated towards Value Demand: 31%
- Overall capacity dedicated towards Failure Demand: 69%
People could relate to what the data was showing. It matched their feelings and anecdotal evidence, and of course, no one was flattered by the current state of their performance. Nevertheless, the culture was forward-thinking, and instead of finger-pointing, they started identifying where the problems were and, for each, started asking how-might-we questions.
After six months, they measured and compared their performance with that original baseline. That’s what they found:
- Lead time (85% Percentile): 58 days or a bit less than two months
- Predictability (in relation to service level expectation): 96%
- WIP Age (85% Percentile): 145 days or a bit less than five months
- Throughput (over last six months): 33 new features
- Overall capacity dedicated towards Value Demand: 65%
- Overall capacity dedicated towards Failure Demand: 35%
My advice: don’t delay the start of your journey. Don’t wait for the utopian perfect state to take the first step. In 24 weeks, just by becoming performance-data-aware and committing to pursue continuous improvement, that program managed to achieve the following outstanding performance improvements:
- shipping software 69% sooner (from 190 down to 58 days)
- 54 percentage points more predictable (from 42% up to 96%)
- Delivering 22% more features (from 27 up to 33 features)
- 34 percentage points more capacity dedicated towards Value Demand: (from 31% up to 65%)
The program has improved its delivery performance and has increased immensely the quality of its data. That included what they were visualising and measuring and the discipline and motivation of keeping the data accurate and up to date.
Before, they were visualising just traditional information like the type of work (epic, user story, bug, task, sub-task), title, assignee and status.
Now, with need driven organically by the data journey, they have pivoted from using standard Jira issue types to using items that represent customer demand:
- Incident & Defect
- Risk & Compliance
- Management & Coordination
- Enhancement & Optimisation
They have also started visualising whether the demand is refutable/irrefutable, delayable/not delayable, the class of service (expedite, standard, fixed date, intangible), and the value area (customer, business, architecture, infrastructure).
This is a classic pattern where looking at the data leads to better performance, which leads to better data, which leads to better performance. It is a reinforcing positive feedback loop.
The report proposes the following steps as part of an implementation roadmap:
Get going from wherever you are.
Start by identifying your value streams. A value stream is anything that delivers a product or service. You're aiming to accelerate the flow of value to the customer.
Find the people accountable for every step in each value stream.
Bring the players in your value stream together for a mapping exercise. Find where the idea starts, and track every step until the value is delivered.
Connect the parts of your DevOps toolchain aligned to the steps in your value stream map and start getting real-time data and insights into your value stream's flow.
You've automated your value stream map, now use it! Set goals for your value stream and use retrospectives to determine where you are.
Use your insights to design and perform experiments that adapt and optimise your flow so you can continually delight your customers.
Set your long term vision and goals.
My experience on the proposed roadmap is that the sequence of steps proposed usually requires a lot of coordination across functional silos, which in large organisations commonly involves friction. This can delay the whole process or, worse, weaken the initiative.
In my experience, the path of least resistance tends to start with “step 5 - Connect” and use key customer-centric and flow-based metrics (time to market, productivity, predictability, quality, flow efficiency). That’s a powerful improvement driver that makes the process of creating awareness, alignment and a joint mission for your value stream Management program more straightforward.
A successful pattern I’ve found includes:
1) Start from where you are now. Don’t delay the start of your journey. Don’t wait for the utopian perfect state to take the first step.
2) Find a Value Stream Management Platform that enables you to measure the effectiveness of your end-to-end product delivery, regardless of the scale and complexity of your delivery processes and tooling ecosystem and connect your work first.
3) Connect your agile boards to the platform (plug and play) and see what the data looks like. Set a baseline.
4) Find insights that can point you to key sources of delay and issues and hopefully give you a hint on possible causes with actionable insights into tackling them.
5) Make some informed and intentional interventions and introduce some necessary systemic feedback loops to see if you need to pivot or persevere.
6) After six to twelve weeks, you should start seeing tangible results with a powerful story emerging and hard data to support your narrative.
7) Share your story with other leaders, accountable for the end-to-end customer Value Stream, as well as from different value stream segments and supporting value streams. Get them on board!
8) Map your end-to-end value stream, identify people accountable for each segment, connect other tools used (PPM, Agile, DevOps, Operations) to get further insights
9) Work together to create lasting changes that will make the value stream more humane, sustainable, efficient and effective.
10) Spread the love, become an example for other value streams.
I already love the work that the VSM Consortium is doing, bringing clarity and guidelines to help advance the industry towards global organisational performance. I can’t wait to see where their research will take us. You can find more about them and download the report here.
I hope the ideas in this content have sparked insights and gave you some food for thought. Click here to experience the Flomatika platform, and If you want to know more about us and how we can help you solve your product delivery challenges, we would love to hear from you!