How data discovery can prevent you from falling in the 85%
Great data projects start from the ground up.
But what constitutes the ground? What are the key foundational steps needed to ensure your data investment pays off as highly as possible?
It's important to ask and understand the implications of this question. Why? Because 85% of data projects fail. This article discusses some core reasons data projects fail, and how we've developed our ways of working to protect clients from falling into this trap by laying the foundations for successful solution implementation.
Let's discuss four primary reasons why almost 9 in 10 data projects fail, and how White Box's systematic ways of working avoid these pitfalls.
Siloed data
Company data is often abundant, and stored across multiple databases and spreadsheets in varying formats, making it challenging to connect the pieces for analytical purposes. Some companies can get away with only using one data source, particularly if they are small. However, the large majority of businesses will have multiple, and auditing, understanding and de-siloing this data is pivotal for laying the foundations of a great data project.
How White Box de-silos data
At White Box, we emphasize the importance of the data discovery phase to our clients. This phase involves identifying all relevant data sources for the project, running them through best practise data auditing techniques, and working collaboratively with our client to communicate the pros and cons of different data access, manipulation, storage and refresh processes. The greater the investment in discovery, the more efficiency we can leverage in the build and implementation phases - a win-win!
Unavailability of skill and resource
A featured article from Towards Data Science by Christopher Zita found that Data Scientist jobs are ranked third best in the entire US job market, yet they take an additional 5 days more than any other occupation on average to get filled. This screams high demand, and low supply in the current job market, warranting high salary requests from those with the skills your operation needs.
White Box acts as your extended team, with full access to senior data scientists, consultants and analysts
At White Box, we work transparently with your team to understand the right solution for you. We then define the project, and the required resources and support necessary for delivery. You then have access to a core project team, as well as the extended team if particular niche skills are required for certain project components. The below diagram explains this system, and the convenience it provides when compared to trying to hire an individual with all of the skills required.
Poor transparency
Aligning expectations between teams and team members is often a challenge. Data projects are often complex, and the goals for the end solution are often measured in different ways. For example, data scientists might focus on the accuracy of their model outputs, whereas business managers will focus on the financial return the model generates following implementation.
We work using a phase based approach, and develop the project scope in collaboration with our client
Transparency is a fundamental value we stand by when building projects. We break down the key phases and associated resources, time and costs necessary to complete the tasks within those phases, and explain the reason for and value in each step. Before commencing the project, we work with our client to build a scope framework to ensure there are no grey areas in the deliverables expected.
Deployment and usability
Data projects often don't consider the end goal for implementation. Without a plan for implementation, solutions can often be left untouched and unused, delivering no business value.
Our discovery phase emphasises usability as a necessity with ongoing support crucial to the success of the project
As already mentioned, investing in discovery pays dividends down the track. One of these dividends is the effective strategic implementation of the developed solution. By truly understanding the business problem, we develop pragmatic deliverables that consider the time, technical skillset and functional desires of our clients. We are also mindful that solutions need to be supported and developed as the business environment and internal goals shift, and offer ongoing support options to cater to these changes. To see some examples of solutions we’ve delivered, feel free to explore our range of case studies.
Interested in kicking off your next successful data project with White Box? Reach out today and we’ll get back to you with the next steps for delivering long-term business value.
Contact the White Box team today
