...Or How I applied it in my renovation project.
tl;dr:
With creativity, applying Agility to Construction is possible. We managed to get access to features early on our renovation project which helped to validate (confirm or infirm) their value and allowed us to enjoy them sooner rather than later, which avoided building up risk for too long. And there's still room for improvement.
Long version:
We recently had our living room and bathroom remodelled and I took upon myself the challenge of trying to apply as much Lean Startup and Agile Principles into this operation as possible.
Sometimes, when I ask dev teams to deliver bite-sized chunks of value, a seasoned developer who's used to spending weeks -- sometimes even months on the same task -- would cite Construction as an example of how it's not always possible to deliver in small increments.
I always find that argument interesting because it is a situation that requires a lot of planning and usually has no margin for trial and error.
When building a bridge such as the Golden Gate, it's hard to imagine how the workers could deliver any value after only one to four weeks of work. Note that I picked that time period on purpose, as that's the definition of a Sprint by Scrum.org.
A product like the Golden Gate is something that is huge and very difficult to change after completion.
Imagine launching into the development of such a product without knowing anything for sure about its value. Luckily that wasn't the case. The need was being answered with ferries or long roads, and the idea that bridges are an improvement on that was far from novel.
Most of that is rarely the case with software. When software is built, it's frequently about coming up with a new solution, not the thousandth replication of a very similar solution.
Software, like its name implies, is soft. It's much simpler to modify after having been written than the Golden Gate was after being built. Iterating on software is far from being unthinkable.
But what about construction work that is not just about building Yet Another Bridge? What about home improvement, where needs vary according to who lives in it?
Regarding our renovation project, our house is quite small for two adults and two young children: only one floor, no basement, one bathroom, and children that absolutely need to go multiple times during the night.
Our living room had an old wood stove that could no longer be used due to stricter municipality rules.
After a family brainstorm, we came up with a long list of values to aim for with this operation (our backlog). Here's a shortened version:
• Have easy access to our coats when leaving the house and quickly be able to put them out of our way when coming in,
• Increase space in the living room while still being able to enjoy a wood fire,
• Have a rustic-inspired decor in our living room,
• Improve toilet availability (not enough offer vs. high demand),
• Improve access to the toilet for people using the swimming pool in the backyard,
• Limit the impact on our lives during construction (we'd remain living there, see budget),
• Maximum budget of $60k.
So the first element that we requested was to limit the number of nights we would be deprived of a toilet. That request alone was a challenge for the constructors, given that the bathroom's entire concrete foundation would have to be demolished, all plumbing redone and a new concrete foundation poured in.
One enterprise said they just couldn't do it in less than 7 nights. The one we hired promised only 1 but started readying us for 4 or 5 nights. They would take one day demolishing, one redoing the plumbing and pouring concrete, and the last day would allow for the toilet to be installed. On top of that, there was a weekend in between, which would add 2 more days.
Here's where the bite-sized chunk of value came into play. Instead of waiting one day to pour concrete everywhere, I asked the senior guy of their team if they couldn't just pour the minimum necessary at the base of the toilet on the same day as the plumbing so that it would already be dry the next day, allowing for the toilet to be installed. To which he replied: "I haven't thought of that. Sure!".
We also had examples of small early deliverables helping us see that the solution initially thought was not good for the value we were after, and the fact that it was a small chunk allowed us to abandon an entire feature and change direction with minimum impact.
One such example was the fixed structure initially planned for storing firewood. It would have consisted of a metal skeleton covered in drywall and finished with decorative rocks and mortar. The metal skeleton was installed bare first. When we got together in the evening and looked at it from the kitchen, we quickly realized that one of its walls would obstruct the view toward the TV. We considered alternatives such as making that wall less deep or less tall and compared the potential visual outcomes against the value of having that storage space. We ended up abandoning that idea entirely and opting toward a movable and/or much lighter structure to be added sometime later.
We would have never had that chance if that feature had not been made available to us in its simplest form so early.
Another parallel with Agile Principles was the frequent communication and collaboration. It helped to avoid nasty problems and quickly resolve blockers when the communication was there, and it ended in disappointment when it wasn't.
In the beginning, circumstances external to the project kept me on site while the work was being done. This allowed the contractors to come to me to confirm details for each feature. I was involved in adjusting the measurements for the walls surrounding the stove and helped to avoid an error of a couple of inches, which would have jeopardized the stove's installation.
Another time, the worker installing the decorative stones told me that he was afraid we wouldn't have enough stones for the job. The stove installer would be there the next day. Delaying the stove was not an option. I quickly discussed with all parties to reach a contingency plan:
• Prioritize the surroundings of the stove so that it could be fully installed the next day,
• Order more stones, to be available within two weeks,
• I'd remove the stove's connection to the chimney when the rocks would be available,
• The Stone installer would then finish the stone job,
• I'd reinstall the stove's connection.
Once again, we had access to a value (the stove) even if the surrounding structure could not be finished before two weeks time. If the communication and the strong desire to prioritize value were not there, the stove wouldn't have been available for months.
There were a few other good hits, but there is also room for improvement in future initiatives.
Some deliverables were delayed until the last minute in an attempt to optimize the work. They ended up becoming examples of early optimizations that get in the way instead of helping.
One of these was a new storage space made from splitting a wardrobe in two. The split was done early, which forced us to empty it early too. But it wasn't until the end that the shelves were installed and the paint job was done.
All that time we had to live with stuff piled up in a room when we already lacked space impacted our quality of life while the feature was not delivered.
Many such features were started in parallel and not delivered until much later.
But all things considered, I'm very pleased that we were able to enjoy pieces of value sooner and I'm happy to see that it is possible to maximize value and minimize risk even in construction work.
When we concentrate on bite-sized chunks of value, it forces each individual to think harder on ways to deliver value in small increments, even if it seems impossible at first.
When applied to our project, it helped us to confirm their value and enjoy them sooner, and to uncover unknowns earlier, which enabled course-correction toward a better finished product.
If you're a constructor that already works in this mindset, let me know. We have other projects waiting!
Image credits:
Others: felipe(AT)carassonet(DOT)org
No comments:
Post a Comment