Estimation game article

Hi,

few months ago I published an article about the agile estimation game we used in project. Beware: it is in German:

https://www.projektmagazin.de/artikel/die-alternative-zum-planning-poker-das-team-estimation-game_1077797

Lean, Agile, Scrum ... and bussines model

After my last post with some content, I got positive feedback from friend, former colleague: "nice post, good idea, I'll try it out". Sure! - Lean-Agile-Scrum are full of great ideas. Too few people are listening to them. Hmmm and then I've got a question: "how could I incorporate it into my business model?". And I could not answer. That's the point => we now already how to build great software, but we do not have any clue what to build.
OK - so I have got myself on the way to learn new area: I have changed books, blogs and activities. Instead of "Succeeding with Agile" by Mike Cohn I read now books from Steven Blank and Eric Ries. Instead of attending big Lean and Agile conferences I attend barcamps and startup weekends. And instead of talking inside the team or organization how to change the process I started to write code again. And making customer surveys, and evaluating ideas, and changing fast my direction. And I nearly quit my job (I spent now 20 hours/week for my employer). Surprise, surprise - my personal life satisfaction went up dramatically. Stay tuned.

Lean Kanban Central Europe - Munich 2011

Cadence and synchronization

Week ago I've finished an excellent book from Don Reinertsen and since then I cannot stop thinking about cadence and synchronization.
Two simple elements.
Analogy: traffic lights.

Cadence
Imagine what would happen if traffic lights were not predictable. Nobody knows when comes next green. What would you do? Some people would slow down on yellow, some would go on red. Chaos. People would start to look for longer alternative, but predictable ways. New source of variance ...

Something similar happens with big products - nobody can predict next release. All are focused to finish actually running one. Some people try to slow down to have it finally finished, some pushes to have "this horrible bug" solved. Or extremely important end-user feature must be in. Many companies, however, already realized the power of cadence: e.g. linux www.ubuntu.com is released every six months. Rock solid for years already. Predictable. Did not your favorite feature managed to get into the release? No panic, no stress, no pressure - in half a year it's in.



Synchronization
Everybody knows this terrible feeling having in one line four times red signal. Solution is simple: green wave.
Development of big products is similar: try to personalize features. How do they feel when after successful passing the design phase they must wait on red signal for weeks to pass through and wait again for weeks to be integrated and again to be tested and again to take over.

Two simple elements: cadence and synchronization.

Neuschwanstein Castle - Lean paradies, part 2

And here another set of my observations after few days off. In between I could participate on great conference ALE 2011 in Berlin - some insights come later.

• Ticket is bound to specific tour
  Clever. This aspect decreases the variance and thus subjective value for visitors. On other side a lot of conflicts can be avoided this way (assignment to certain tour takes place already when buying ticket)

• The number of tourists in one batch/tour is limited
  Simple rule to decrease variance and thus increase predictability. Whole tour is then optimized to this number in one batch.

• You cannot buy ticket for a tour that is already sold out
  Relevant to previous point - however it says that information about maximum and actual batch size is available already on ticket counter.

• In front of the tour entrance you can find people in three queues
  - prepared as one batch (5 minutes to start)
  -  finishing the collection of the group for tour (10 minutes to start)
  -  just opened queue (15 minutes to start)
  -  the rest of visitors is waiting around until their tour is next to be collected in waiting queue 

• Enter to the tour queue is automated based on the ticket (to minimize transaction costs)

• Last stage in the tour is without explicit guide - The guider is already on the way to next group that is in-between collected in the first room of the castle

• No tourist gadgets can be found in the entrance area - these are side artifacts that could influence the throughput

• During the tour you can hear small bell. I am not sure, but I suspect that this is a synchronization signal for all tours inside the castle. Every guider gets a feedback if his group is too slow or in opposite - should slow down. This keeps the process cadence stable during the whole day.

• Whole capacity is not used on 100% to keep the overall throughput on maximal level. (See Allen-Cunneen Approximation for queue length and resulting process performance as consequence of Little's Formula)

• And results?
  -  People are happy as service is excellent
  -  More than 1.3 million people visit annually, with up to 6,000 per day in the summer. This means up 10 hours every five minutes one batch with 25 people, with very small buffer for unpredictable events. Wow.

Neuschwanstein Castle - Lean paradies, part 1

Recently I had the chance to visit castle Neuschwanstein. And surprise: beside not discussable cultural value I could observe many lean practices.

I cannot comment if all of them were implemented on purpose and under the Lean umbrella - I have not noticed explicit lean logos as we know it from Toyota.

To save some words here is the inventory of my observations:

• Ticket sale service is in separate building – not in front of castle entrance
  -         This is queue content optimization pure :). It provides the flexibility to re-organize content of the big queue so people are approaching the castle itself in already given order (sure with some variance –  respecting individual needs)

• Ticket sale service is equipped with lines for people queue
  - This technique is well known and can be often seen on airports when waiting for check-in. This tool ensures that incoming order is clear and predictable. A lot of conflicts can be avoided.

• The length of people queue is configurable in ticket sale service. People queue itself is very effective, but it could slow down throughput if the installed length does not correlate with real needs.

• One queue - many ticket counters. This way, in this case, is more effective than the other alternative: every ticket counter has separated waiting queue (can be seen in supermarkets).
  -         One can increase the performance of this process step (selling tickets) without need to reconfigure the queue itself.
  -         Ticket counter that unplanned stops working does not increase blood pressure of 20 people waiting exactly in this queue. Both this properties are very valuable in high

• Tours are guided only
  -         This is core element, means whole process is under certain tact, cadence.

Impact of visualization on the self-organizing system

In self-organization systems/teams, visualization plays an important role. Consider next graphic expressing the definition of project structure.

 

In the project organization structure two extremes can be observed: chaos or command and control.

Chaos in project can be seen, e.g. if team structure is not clear, every co-worker is (maybe) working on topics he founds important. This leads to local optimizations on different levels. The probability to succeed in this structure is very low as can be often seen already by first integration of outputs.

The other extreme is the “Command and control” approach were single person – typically technically very strong – is driving every single change personally. He gives orders and on regular manner checks the progress, takes countermeasures if necessary, etc.

Within scrum – what we want to achieve – is such level of self-organization, that pros from both extremes are in place and if possible cons are eliminated. Visualization can help establish so-called center of gravity – a motivation point – attracting people and driving them as a team towards common goal. It is the common goal that drives self-organizing group to the right direction and help them consistently work to achieve results.

As only very small portion of  available information can be physically visualized, people are very instinctively attracted.

Similar effect can be caused or is supported by good vision.

On the other side it could be also dangerous tool. As only small portion of knowledge is visualized, it is important to show correct part of it. However, I believe in healthy and critic environment - folks discovers “side ambitions” very quickly and pushes company towards open and transparent culture.