For the second year in a row, I was fortunate to be able to attend Devjam‘s annual Agile Day Chicago, on November 4th, and the theme was Anti-Fragile.
This year, the company that I work for, Centro, was able to host a couple of “bookend” events for the main event – a CAOS Meetup with a panel discussion on Anti-Fragility a few weeks before the event, and the fun Beer-o-spective/Happy Hour after the event.
The day started with traffic being snarled as people arrived in droves downtown to celebrate along the Cubs parade route (Go Cubs!), and fortunately, the conference still had a great turnout. Approximately 300 gathered in Morningstar’s offices, and the full agenda/program can be found here.
The theme of the conference is based on the book Antifragile by Nassim Nicholas Taleb, and in the opening remarks David shared that the main takeaways for us to keep in mind were:
- The difference between the level of fragility of a wine goblet versus a washing machine.
- How the human body can contract a virus, create an anti-virus and evolve.
- To constantly challenge the status quo.
- To look for ways to evolve (make less fragile) the things we can control (terrible code, solvable problems, move past current situation) versus things we can’t control, like the market conditions.
He also brought up the work that Netflix has done with FIT (Failure Injection Testing) as an example that we can all learn from for how to test the fragility of our systems and applications.
There were three sessions to choose from each of the three timeslots, and the I chose to attend Shift from Doing Agile to Thinking / Being Agile, which was about Morningstar’s journey from “simply following the protocols of agile to actually living the principles.”
Jason Stipp and Maria Hamidi provided a great overview of how their teams were set up, roles and responsibilities, how they live the principles, their ceremonies, process and learnings.
Once the presentation is available for download, I’ll make it available here.
What I enjoyed was learning from another organization the pain points that they had and how they approached solving them with what works best for them. I really appreciated how they articulated the roles and responsibilities and plan to leverage some of the content (as they say, imitation is the…).
It was a lively session, with lots of questions in a standing room only room. So many questions and discussion that we ran out of time and had to skip some of the details, so the presentation slides will come in handy!
The highlight of the conference for me was being able to see Mary Poppendieck speak in person. Her topic was Platform Beat Products, and it focused on how companies that provide platforms (think AirBNB, Uber, Amazon, TripAdvisor) are excelling in the current markets.
She began with how the markets have evolved from the traditional pipeline of R&D -> Raw Materials -> Manufacturing -> Marketing Sales -> Customers to one of creating a platform for communities to connect and conduct business – helping people share value exchange. Gone are the days of needing Travel Agents and Phone Operators.
Core value exchange involves two parties and makes it easier to do business, and every platform needs to identify that core.
The platform also has the find the friction. Mary told the story of Malta, Greece, a small island where there isn’t any vehicular traffic, so when goods (for example, a flat screen tv) is dropped off by boat at the pier – how does one get it home? The secret isn’t to find the features, it’s to find the friction.
Next, Mary explains that the Platform has to answer the question of how they help the community. She shared the story of AirBNB who have Community Managers in all the cities they’re in, and the Community Manager is responsible for advocating for the Owner and helping them find people to rent their homes.
For example, one Community Manager came up with the idea to hire (and pay for) professional photographers for Owners so that they would have great photos of their homes, which makes a renter more likely to choose a home.
Next, the Platform needs to ensure that the environment/community is “fair, comfortable and enjoyable for all sides.” They also need to be the ones to “make the rules” to ensure that outcome is created.
Then, the Platform needs to be able to scale – for example, AirBNB doesn’t need to invest in building new hotels like Mariott does.
Below is the wrap-up of how to succeed with Platforms:
For the final presentation, I chose David Hussman’s Product Agility 0.1.0, which is an evolution over the previous mindset of just Product Over Process, it includes Scaled Product Learning and Learning Dojos.
The concept is to shift the current production model to one of a learning model – are we building the right thing? It’s about product, people and validating ideas, and looking at how wrong we were when we didn’t build the right thing.
David explains that in the 90s, it’s was about the Project (on-time, under budget, with good quality), then in the 2000s, it was about the Process (but we’re doing daily standups!) and now it’s about the Product.
He explains that these are on the continuum of the levels of Certainty to Uncertainty. When we were really certain, it was more important to build the “thing” right (project) to our current level of market uncertainty – we need the product agility to learn what the right thing is to build.
He then stresses the importance of team ownership over product ownership (something I completely agree with and why we decided to call the cross-functional, product development teams at Weight Watchers Product Ownership Teams – yes, POTs.)
He also begins to overlap the concept of Discovery and Delivery (where previously these were considered different phases), and in the overlapping part, explaining the importance of storytellers that can speak to why the product/feature is important to the customer’s journey or story – the why, who, what and where. He calls this Early Discovery.
He also shares a concept of adjusting the production amount to allow for additional discovery when needed.
In the spirit of “you get what you measure,” he suggests that teams measure impact vs. progress and ask questions regarding the value of the feature and are there ways that we can apply test-driven development practices to user experience – don’t start coding until we can measure impact.
Then, there were a few riffs on the Agile Manifesto that are always fun:
David admits that he can see this working really well with one product, one team and one piece of tech, and there are many questions when it scales out to one product, multiple teams, or even multiple products and multiple teams. I agree that the answer will be less about process and more about information flow.
I’m curious how the Product Agility movement will grow and would like to be a part of shaping it as much as possible.
Afternoon – Lunch and Open Space Sessions
At lunch, it was great to catch up with folks I hadn’t seen in a while. Then it was time for the Open Marketplace, where conference participants suggest topics for Open Space Sessions – free-form conversations about whatever topic is selected. In a lively session, the board filled up quickly:
I choose to stay in room 10 because all the topics interested me. The first hour were deeper dives with David into Product Agility, and then it was with Mary Poppendiek on No Backlogs and No Estimates.
In the deeper dives with David, we started to talk about the concept of a Definition of Ready for a feature to move into delivery (does it have the right level of validation) and the concept of Discovery Cadance and the importance of the team being involved (the “what abouts”/ critical thinkers are crucial at this stage).
There was discussion on:
- the importance of validators that are not the user
- that ideas can be submitted by anyone
- the distinction between a Hypothesis and a Guess (a Hypothesis is an informed Guess)
- the importance of Story Maps and Customer Journeys (where do you want to take this person)
- trying to figure out the level of visual design needed and what the design dimension of the stories is; including a friendly debate between Mary and David on the concept of a Design Sprint.
- the concept of “appropriate fidelity” – not just Hi or Low.
My big takeaways from where David is going with Product Agility are:
- Challenge the assumption that the Product Owner / Product Manager is the only person that can add ideas to the roadmap/backlog/etc.
- Challenge the assumption that all items are validated to the current market/customer need.
- For items that are not ready/validated – what’s the line for “good enough” or satisficing.
- How do we scale product learning to help create more validators and idea submitters?
Next was Mary Poppendiek explaining her “No Backlogs” concept, which was one that I proposed because I’m always interested in hearing the controversial side of an argument.
What Mary explained was the importance of work in progress limits to create just in time manufacturing. Backlogs are evil because they cause “inventory” to “pile up on the manufacturing floor.”
Even item is a demand generator, and it will always make someone wait. Taking the calculation from Lean where Time Spent Processing / Time Someone is Waiting for Work Solution to begin to show Flow Efficiency.
She stresses the importance of understanding strategy, team capacity and filter demand based on those constraints so that there is more intelligent demand created. This means saying “no” to requestors versus just adding their request to a backlog that is already miles long.
For example, if a team knows that “small” items take two or four weeks to deliver, they should only fill their “buffer” with a few small items that will allow them to tell the requester that their request will take six weeks (assuming items wait in buffer for two weeks).
I can see the merits of the concept and the process/operational person in me says this makes perfect sense. The team using it would need to be very informed in their capacity and ruthless with saying “no” to requesters (or maybe not now, but come back later), which could ruffle feathers (but then again, just saying yes, adding it to a backlog and not picking it up for years isn’t a great alternative either…)
Because I needed to make sure the Happy Hour was off to a good start, I wasn’t as engaged in the No Estimates as I would have liked, which turned into a conversation about how consulting companies could leverage no estimates. It was interesting, but after hearing Woody Zuill’s presentation at Agile2016, the conversation was heading towards the literal discussion vs. the sizing discussion that he encouraged.
Beer-o-spective Happy Hour
Despite many conflicting Cubs celebrations, we had a nice turnout for the Happy Hour at Centro. It was great to keep the conversations and networking going with a few dozen people in our beautiful office space.
Looking forward to Agile Day Chicago 2017!!