How to Rescue a Failing Data Project in 7 Proven Powerful Ways
A failing data project is as certain as death and taxes.
I recently came across the 2mm rule by Tony Robbins; the gist is that small changes can significantly impact your trajectory for success. Throughout my career, I have seen many failing data projects. Most had the writing on the wall, but they continued on the path to doom for various reasons.
Today, let's learn how to rescue these projects from total failure.
1. Understand the Current State of the Project and Gauge How Far It Is from the Desired Outcome
The first question to ask is, what is the desired outcome?
It sounds simple, but you'll be surprised to hear how many projects can't see the wood for the trees. The desired outcome could be a report, dashboard, a predictive model helping with a business decision. The desired result won't be a data pipeline, data model or data warehouse; they're a means to an end.
Impact on Data Teams
Data Teams need to start looking at the problem holistically. Data engineers can't just focus on their pipelines, but what is the end impact on the business decision due to a poorly designed pipeline? For example, the Finance teams cannot pay commissions on time due to poor performance of the data pipeline.
2. Engage with the End Users and Negotiate an MVP (Minimum Viable Product) of an MVP
Nowadays, projects tend to have a defined MVP, but what is the MVP of MVP?
There is a psychological barrier to project delivery each time an issue arises which deviates from the end goal. The end users' requirements may be loose, or the technical debt pile may be too high. Allow the project teams and the end users to see the light at the end of the tunnel by breaking down the project into smaller components of meaningful delivery.