Justified for Taking a Lean Approach?
On a project I’m currently involved in the client found flaws in their proposed system very quickly thanks to my team applying lean concepts such as building MVPs. In this post I’ll share our story so you can achieve similar successes in your own career.
An interesting theme to this story is that I am new to this company and so are lean concepts. I had to deal with team friction to lean concepts without burning bridges along the way to helping the client achieve their goals.Tell Me About This Project
Our client want to bring their comparison platform up-to date whilst maintaining their place in the market as offering the cheapest deals available. As I blogged about previously there is a big design trade-off between the speed at which results can be shown and the ability to always show the best results.
The main problem is that each search request hits multiple suppliers who all respond at different times. It’s hard to apply business rules to a changing data set and give users the best experience.
But it wasn’t until we had a working prototype that the client could discern this…Being Lean in a Waterfall World
At the start of the project the project-management team had the traditional phases laid out — R&D, design, build, QA, deploy. Yet I could see ambiguities and holes in the requirements. Unfortunately there was little mention of getting early/constant feedback from clients.
I had to be subtle and respect my colleagues if I wanted to be lean and increase agility on this project. Instead of rushing in I chose to be opportunistic by capitalising on small opportunities….
The R&D phase was to establish a baseline for some technology choices. We created a prototype to achieve this and quickly agreed on the technologies. I took this opportunity by making the prototype available to the client, with team agreement, to get some early feedback about the proposed functionality.Internal Friction About Low-Quality MVPs
As we left the R&D phase there were still unkonwn functional requirements so I viewed the prototype as an MVP. I encouraged, and led, the team to add un-tested, half-baked features with just enough quality to let us and the client see how the system would work if we built a full version that way.
From the client’s point of view this was great. We had deep sessions walking through the prototype, iterating rapidly on concepts — because they were low-quality we only spent the minimum amount of time necessary building them which allowed us to iterate quicker.
Internally it was a different matter. I felt I was creating a reputation for jumping around, only doing half a job here and half a job there. A few comments suggested that long-term architecture and direction were required.
In the face of uncertain requirements, I believe those activities would have been wasteful.
To resolve these situations I used facts politely such as: “we don’t know what the client wants from feature A. Feature A impacts upon Feature B, C, and D. So we can’t build or design a system because most of it is subject to radical change.”.
With their differing beliefs the team, som grudgingly, accepted this view although small elements of friction did remain. Luckily the people I work with are exceptional characters, so any disagreements never became personal. We still had lots of laughs and maintained team harmony. I fully encourage you to always try and maintain team harmony.Then the Client Made a Pivot
After a meeting with the client they really started to understand incompatibilities between the proposed federated search, their business model and the user experience they wanted to provide, thanks to us walking through the latest MVP together.
The next time I met with them they had decided to change aspects of their business model that would allow them to have more ownership and control over the data. This would eliminate the federated search allowing them to fully express their business philosophy. I don’t know for definite, but I expect without the MVP this process of discovery may have taken another 3 months.
I think that justifies the lean approach.
