Agile Roundtable Notes September 2010

The Salt Lake Agile Roundtable meets the first Thursday of every month at the Borders bookstore in Murray, UT from 2pm to 5pm. These are my notes from the September, 2010 meeting.

Evan talked about the packing peanuts exercise on tastycupcakes.

Robert brought up the idea of opinion-driven development. When technical considerations are not the driving factor, is there a way that you can use the agile values to guide the decision making process instead of just devolving into a battle of opinions? Zhon mentioned that Jeff Patton’s story mapping method can be used to guide discussions and decisions. This wandered off into a discussion of technical debt and the software craftsmanship movement. Getting distracted by arguments of opinion is a waste of time; when pairing Zhon will often shortcut the argument by saying “let’s try yours”. If their idea works and is good enough, then the roadblock is removed and we can continue making software. The goal is to get the software shipped as soon as possible to validate the hypothesis that the constructed software is wanted in the marketplace.

Donna Harris has a bookkeeping service and is interested in agile marketing to improve her business. She heard about the roundtable from last month’s discussion of agile marketing. Referrals have been the most effective method so far. Nate’s company doesn’t use any marketing that can’t be tracked for effectiveness. For instance, when putting your phone number on a web site, put a specific phone number that is only listed on the web site. Then after a month when the phone bill arrives, you can identify the effectiveness of listing the phone number on the web site. The agile approach is to get information based on experiments in order to close the feedback loop and get a better idea of what to try next. It was suggested that if your customer sends invoices to their clients and your company is preparing the invoices, then you could add free word-of-mouth advertising on the invoices, similar to the way email messages are tagged “sent from my iPhone” as a word-of-mouth advertisement for the iPhone. Donne is going to try out the idea of making video testimonials and is going to start asking every small business she interacts with “Who does your books?” as a way to build contacts.

Mike Moore talked about a conference he is organizing and is interested in some ideas for how to make the conference better. What kind of collaborative exercises do you like to participate in? What is the best way to highlight this conference and its collaborative nature? This is not a sit in a chair style conference, its very hands-on. Robert suggested getting a schedule up on the web site. Mike wants to make sure there is time to recover between sessions by providing a schedule without sessions booked back-to-back. Nate suggested having a rails bootcamp track to bring new ruby developers into the mix.

Evan wanted to explore the idea of treating all the projects in your company as one big agile project. Organize the projects into sprints at the global level. Has anyone seen this type of thing and what were the experiences? Zhon said that projects should be allocated at the 100% level; don’t split resources 25% between 4 projects. This makes sure that any project you try is sufficiently allocated to have a chance of success. This is similar to completing stories instead of just doing half the effort for a story. Andrew raised the point that this approach tries to identify which projects should be abandoned. Do we keep this project or do we discard it? Robert tied this back to the idea of economics: how often can a customer tolerate a release? Can releases be used to quantify which projects to keep? How are you measuring which products are performing? When you switch priorities can you really switch teams from one project to another? Are the teams too tightly coupled to the products on on which they have historically worked? Some engineers like the historical association and some engineers like to change up what products they work on. Zhon says that CEOs have two jobs: buy and sell companies and be the face of their company outside. Remind CEOs that this is their job. Andrew disagrees with Zhon’s premise and believes that the culture of a company is the job of a CEO; Zhon believes that this is reflected through being the public face of the company. Are there zombie projects that are just continuing even though they’re really dead? Jonathan wants to be the killer of zombie projects.

Bill is starting a company with social media marketing. He is interested in applying agile processes towards his marketing efforts that he provides for his customers. Robert mentioned that there is a 70 employee company called SwarmBuilder in SLC that is doing marketing that eliminates the middleman.

Kay wants to talk about certification. Would the certification for agile developer imply a certification for developer? Or is it just a certification for the agile part? Does it imply that they have a certain level of competency as a developer? Alistair says that an agile developer should be able to do things like slice a story down into finer and finer chunks. Alistair thinks that the agile skills are overlapping with developer skills but not a superset and more closely like a subset. Alistair thinks that the core of an agile developer certification is focusing on the agile skills and not on the developer skills; placing more emphasis on delivering the most business value soonest as opposed to specific programming skills. Alistair doesn’t like on/off certifications, he likes levels of certification — shu, ha, ri again? Alistair doesn’t think so.

I work with a non-profit organization that responsibly recycles computers and refurbishes them for low income families and non-profit organizations. Like other volunteer organizations I’ve worked with, they have problems controlling the chaos of their volunteers and harnessing their energies in the most productive fashion. I asked the group if there were any techniques or practices from agile teams that could be applied towards the organization and the problems they face in effectively have in harnessing their volunteers. It was sugested that we could create information radiators to decorate the space and clearly indicate the most valuable things that volunteers could work on. We could use an iceberg list to bring the most valuable tasks to the top. An iceberg chart always has the most important thing on top and you always pick the highest thing on the list that you can accomplish. The most important thing is to organize these radiators so that the work is self-organizing to avoid having to micromanage and coordinate all the activities. Another suggestion was to make a two minute orientation video for each station in the disassembly/recycling line or processing line. By making things more visible and ready to work on, it will invite volunteers to participate; create a “conformative workspace” . Create a physical token for the work need to be done. You take it to your workstation while you’re working and return the token when you’re done. This creates a visible information radiator for the work that other volunteers are doing to avoid duplicated efforts when a large number of volunteers show up.

Zhon has some “aha!”s from Agile 2010. James Goebel of Menlo Innovations: If you’re willing to make a $40K mistake do one-month iterations. We make $10K mistakes with one-week iterations. 60% of the projecct cost is development, the rest is management, marketing, etc. They turn clients away if they think the project won’t be successful. They do stock trades with companies that can’t afford their rates. Taking stock in a customer’s company is not a big revenue generator yet, but they are still doing this practice. They assign pairs; people like it because they don’t have to pick the pairs. Pairs are assigned for a week. Its not an autocratic decision process, they take into account personalities and strengths/weaknesses when deciding pairing. When two people don’t like each other, they purposefully assign them to code that neither person knows so that neither has an advantage over the other. Client can only sign up for 32 hrs/week. Customer has to pick what the team works on every week, or the customer is on hold. Its been running for 10 years and has been making this system work. Andrew finds the quality of their developers impressive. At AgileRoots 2010, Zhon’s java unit testing kata was only completed by the team from Menlo in the available time. Each person from Menlo finished the kata on time working on it separately. They are just very practiced at doing code in this fashion.

Zhon presented a session on the agile roundtable at Agile 2010. About 10 people came to their session, so they just ran the session as a roundtable. One guy wanted to leave, but stayed due to guilt because it was such a small group, but in the end was really glad that he stayed. Zhon gave meta observations on what just happened around the table. Zhon had someone else facilitating so he could show how to debug the facilitation. For user groups, this could take the pressure off the leader and put the responsibility back on the group. A roundtable was considered exponentially more powerful for solving problems in a work session. One person expressed amazement that they got so many new ideas in such a short period of time, which had never happened before.

Quotables: “Do you have a business card?” “No.” “Fail.”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: