How are betting tables cadenced, e.g., one every two weeks and another every 6 weeks in parallel, based on trigger point replenishment, i.e., when a single queue of pitch submissions is reached, etc.?
We are 3 people on the betting table and 2 of us are shapers too
We tend to do it on Cool Down period.
We have cycles of 4 weeks + 2 of cooldown.
So every 4 weeks we meet to discuss the pitches.
¿What’s your cadence?
¿What did trigger your question?
You time you betting tables? How do you do that predictably when random fluctuations in demand between 2 and 6-week bets arrive at the table?
We make a difference between betting table for “Strategic Projects” thinks we want solve but are not urgent this things can wait until we gather again.
Reactive work which is mostly on demand and comes up when you don’t expect. We make some capacity for this kind of work we have a team that alternates on every cycle that attacks this issues. This is a list that gets triaged on demand (but asynchronous) per week.
I don’t know if I’ve answered the question
Between each cycle, during the cool-down, to set up the next cycle. Similar to what the book describes.
So, betting tables arrive at different times since not all cycles begin and end at the same time?
All our cycles begin and end at the same time.
If a project has an appetite of 2 weeks (small batch) we allocate two of those so it would complete a long batch and everyone end at the same time. (Our long cycles are of 4 weeks, so we have 4 weeks + 2 of cooldown).
I think the rhythm and cadence that everyone starts and ends at the same moment is important.
In queue theory if someone finish always early it generates overhead to the next phase and demand more from the phase before.
But not all cycles are the same length.
I think the thing that is missing in the literature is this: Betting tables happen every six weeks (with overlapping cool-down periods for two weeks). Any given year would begin with a betting table. At any given betting table, if a team is to work on a small batch (two-week cycle), they must decide on two more consecutive small batches to ensure all teams (the 6-week cycle and the 3 2-week cycles) are meeting at the next betting table at the same time.
There are 50 weeks of this out-of-cycle and in-cycle work per year (a total of 6 six-week build cycles and 7 cool-down periods per team). There are always 2 weeks at the end of the year left. A company must decide what to do with those 2 weeks. Perhaps a good way to mix things up is to design a hackathon for the entire company at this time.
These frequent and predicable betting table reviews help synchronize arrival to capacity, which keeps queues short, or removes them altogether, and, most importantly, ensures what is processed through them are of high value when engaging in short-horizon planning. @rjs Agree?
Sounds like there may be some confusion around the way betting table works with smaller batch projects? It also doesn’t help that we often talk about X week cycles, but then always have to caveat with the +Y cooldown weeks. Thus betting tables actually happen every X+Y weeks.
Here’s how we do things:
Our cycles are always 6 weeks (+1 cooldown)
Therefore our betting table happens every 7 weeks (during the cooldown week)
At the betting table we’ll look at what teams we have available over the course of the next 6 weeks and what pitches we might want to take forward.
Pitches may vary in length of appetite from 1-2 weeks or 6 week (small batch or big batch) - teams who will be working on small batch projects are given enough small batch projects to fill the 6 week cycle.
From The Betting Table | Shape Up
“These teams will either spend the entire cycle working on one project, or they’ll work on multiple smaller projects during the cycle. We call the team that spends the cycle doing one project the
big batchteam and the team working on a set of smaller projects the
small batchteam. Small batch projects usually run one or two weeks each. Small batch projects aren’t scheduled individually. It’s up to the small batch team to figure out how to juggle the work so they all ship before the end of the cycle.”
When it comes to planning out the year we’ve found we don’t really need to overthink this, start your cycle whenever you want and plan to adapt throughout the year. Since we’ve just come out of Christmas holidays we had an extended cool-down period, you could also choose to do a smaller cycle of 4 weeks or so where you only commit to doing small batch projects. Since cycle + cooldown takes 7 weeks for us we could calculate that we get 7 / 52 weeks = 7.42 cycles per year, but it’s not something that we worry about.