I don't get excited about a lot of talks, and even less by non-engineering talks, but by the middle of this one I was ready to nominate Susan Rust for president! I didn't catch the talk at Drupalcon Portland, but I did see this session presented at QuaDrupal, a Drupal event hosted by Qualcomm.
Velocity & Viscosity: Train-Wrecks and Ugly-Baby Meetings (From Drupalcon, 2013)
Susan light-heartedly articulates reasons projects fail, and more importantly, strategies to avoid project failure.
There is a lot of truth in Susan's list of reasons projects fail, but my favorite called out was Design over architecture.
Susan explains how putting design before architecture can negatively impact a project through real-world examples and great analogies such as "Build me the draperies before the windows" and "Focus on trees, not the forest". The priorities should rather be analysis, engineering, then design.
Other reasons for failure include clients with no time, bad project match for your company, lack of client trust, and an un-coachable client.
The last one is a zinger. The whole idea here is that the client is not an expert in our field, and if they are not receptive to coaching, the project is more apt to fail, cost magnitudes more than it should, and make your life hell as the producing shop owner or engineer. Of course the producing shop needs onboard for the client coaching to work. Susan's strategies for success are not just for clients.
Susan talks about "Client Wrangling", which is about working with the client very early on to force a project into a Minimum Viable Project (MVP). This sets a project up to succeed by making it possible to establish clear goals and measurable milestones. I believe achieving a significant difference in the MVP vs a full feature requirements is a goal and not something you can just slap on every project. In the cases of porting sites from one platform to another, or a rescue project, I don't there there will generally be a huge difference in MVP vs the kitchen sink, but there certainly could be for a brand new site.
I'm straying from my expertise in analyzing the project management side of things, so back to what I know.
As an engineer, what I see the most value in Susan's project success strategies are:
- Impact, or even that there is a possibility of impact, in their requests
- Yes, the engineer can do that but there is impact, cost, and technical debt to consider
- Concept of technical debt
- $50.00/hr quoted from shop A is likely not an hour to hour equivalent when comparing shop B's quote $150.00/hr
- Clients need to put in a lot of time into the project also
- Clients have greater chance for success at a lower cost if they are coachable (requires trust)
- Architecture before design
- Concept of technical debt
- Set realistic expectations and understandings with clients up front
- How to write a task/issue ticket (Huge peeve, I've probably made enemies over this one)
- A coachable client mitigates risk
- Susan's strategies and project organization will benefit your shop even more than your clients
I may have missed details during the last quarter of the talk as I was busy fantasizing about a world where I would only take on projects where both the shop and client had gone through a Rust Client Certification program.
The reality is that it takes work and commitment to make changes in processes (including a lack of processes), but if any of this resonates as a potential solution to project problems, the efforts are likely to result in lowering risk and cost while raising customer satisfaction, project success rates, and internal morale.
There was more time spent on details such as percentages of time spent on discovery & planning at the Qualcomm talk, which aren't in the Drupalcon talk. There was a video camera running at QuaDrupal, so I'll post it when it surfaces. In the mean time, you can watch the Drupalcon video.
Susan is available for consulting and Client Wrangling. You can contact her via the profile linked below.