Creating great work requires a great Team, and a great Team doesn’t happen by accident or mysterious alignment of the stars. Make it happen by running this checklist at a minimum at every Standup to make sure your Team is ready for the day at hand;
- Epic and Critical Number - on schedule?
- Backlog - 2 weeks ready of backlog, not more than 4, and nothing else?
- Validations - have all Pre-Do-Post Validations for all applicable cards been done?
- Self Contained - no cross team dependencies?
- Protected - requests are queued, not interrupt driven?
- Visible - all work is accounted for, nothing hidden?
- Meetings - only pigs, quick and efficient, grouped together, not scattered across the day?
- Work In Process - no one has more than 2 cards across the active board?
- Stuck - no card has been open for more than 2 business days?
- Swarming - collaborative ownership of cards rather than individual effort?
If the answer is no to any of the above what corrective action are we taking?
It starts with every member of the Team knowing exactly where the Team is going and why? Everyone understands the problem to be solved and why they need to solve it. And they know how they can measure their progress along the way. They know what the one Critical Number is that they can use to measure their progress towards delivering the Epic they are working on.
To achieve the Critical Number the Team needs to have work that is ready for them to do. Work that is defined just enough to allow the Team Member to be creative about how they solve it, but not too little as to leave them guessing as to why they need to do it or what the outcome of the work should be. And they need an ample supply of work to do just in case they get it done faster than expected, but not too much work such that it could become stale inventory and have to be redefined again given the latest lessons learned from the validations performed.
All work on the board has to be immediately actionable within the next 4 weeks. Anything else on the board is clutter and it shouldn’t be used as repository for to do lists, or might do someday backlog, all of that is noise. Get it off the board. You won’t be able to see the forest for the trees if you don’t.
Pre-Validation; Ideally every card on the board has a problem to be solved and a measure of success that we can use to determine if the card achieved the outcome it was intended to. Before a card can move into our backlog it has to be Pre-Validated by the Team to confirm that it makes sense, the measure of success is achievable, and the points for it have been agreed to. If a card is a quick fix, or part of an overall Epic that it is in service of and that Epic has a Problem To Be Solved and Measure Of Success, the time required to define and validate it isn’t always warranted. This is a Team judgement call.
Do-Validation; What we decided we should do a few weeks ago may no longer be accurate or relevant based upon what we have learned recently. Limiting our backlog to less than 4 weeks of work helps lessen how often this happens but it doesn’t prevent it. If our cards are kept to small iterations that can be done in less than a day, we have 3 to 6 people on the team who are all shipping cards, and every card has a validation on it - a learning component, it is very likely that at least 1 card is going be validated that causes us to rethink the assumptions for something we thought we should a few weeks ago.
To make sure that nothing has changed or been learned that could invalidate the assumptions for a card in the backlog the person who picks up a card to work on should at the same time grab their next card, make sure they understand it completely and that it makes sense to them, and they should add to the card how they are going to address the problem. What their implementation strategy will be. Once confirmed and updated the card is put up to the team to verify their understanding and implementation plan. This pre-validation of just in time backlog inventory should be happening on a constant basis to confirm that the work to be done is still meaningful.
Ship your current card > Pick the card that you pre-validated last > Pick another card from the backlog and queue it for pre-validation > And repeat.
Post-Validation; All Epics, and cards that stand alone of epics, once completed must be validated to determine if they achieved the outcome that they were intended to. Did they produce the measure of success that we wanted in terms of one of our key performance indicators. Did we move the needle for that KPI in the direction, and the distance, that we said it would go? Failure to achieve the measure of success that we wanted isn’t a problem. Failure to take the time to learn from the validation and determine next steps to continually improve is. No validation means we have no build, measure, learn cycle, which means we are repeating the same old every day and not improving. Validation takes work, it needs to have points and it needs to be queued in the backlog. It is not an afterthought that we quickly discuss during retrospection. It is the most valuable work that we can do.
To be as efficient as possible the Team needs to have full control of the work to be done. The Team needs to be self contained and have no outside dependencies wherever possible. Everything they need to ship the Epic resides within the 4 virtual walls of their Team.
The Team needs to be protected to focus on the work at hand. Every time a knowledge worker is pulled from the “flow” of completing their work it costs at least 45 minutes for them to return to the level of productivity that they left off at. Every distraction. Every time. If you have 8 distractions in a day and each one causes a 45 minute fall from “flow” it is easy to understand that out of an 8 hour day you could be dealing with 6 very unproductive hours.
There are no hidden forces draining productive hours from your Team - no non-visible work that is being done by Team members that isn’t on the board. This rogue work sucks the capacity out of the team. These hidden distractions will reduce focus and velocity and this invisible work has no validation and learning component to it. It is a rogue wandering to do list of some sort or other. And, the team, and the rest of the company for that matter, have no visibility on the work which means that someone, somewhere, is going to be in for a surprise, of some sort. Either kick the person who is serving other requirements off the team so that they can join whatever team has this mandate, or, make them a chicken so that they can observe but not distract, or, bring them back on as a pig and get visibility and meaning to that rogue work.
Meetings are the bane of the modern workforce. Too few and work happens but it might have nothing to do with solving the problem at hand. Too many useless gatherings and nothing gets done because everyone is in a “meeting”. To find the happy median make sure meetings have a checklist, an agenda for what must be accomplished. Only those that are contributing to checking off the checklist boxes should be there. Meetings need to be concise and focussed. Leave working out the details to those who own those problems, don’t make it problem solving by committee. Schedule meetings at the start or end of the day, and wherever possible back-to-back. Don’t scatter the interruptions across the day.
Every card in addition to the card in motion on the Scrum board, is a source of distraction. This includes cards that are sitting in blocked as nothing in blocked remains quiet for long. There are follow-up questions and ongoing discussions to try and clear that blocked card. Blocked cards are noisy. There are cards up for pre-validation and there are always clarifying questions to be asked on those cards. More distractions. There are the cards that are in play with the other members of the team and their questions and pull request peer reviews. More distractions. Never have more than two cards in play per Team member across the active board, which includes pre-validation and blocked. If you exceed this work in process threshold have the Team swarm to clear the cards. Don’t add anything else into the mix until you drop well below your threshold.
Suffering in silence makes for a good martyr story but does nothing for hitting a Critical Number and closing an Epic. If a card is open for more than 2 days and the owner or owners are having trouble, swarm. Have the Team jump into help. Many hands and minds make for light work and amazing problem solving. If you are stuck, sound the alarm.
If you have 4 people on the Team all working on different cards you are having 4 different conversations throughout the day. Whereas if you have 4 people working on the same card, if doable, not always, but in an ideal world it can be done, then you are having 1 conversation. There are no distractions. And aside from the distractions there is the benefit of having more than 1 mind at the task at hand. More options are presented and discussed, more “eyes” are double checking to make sure that nothing is missed, etc. etc.