During almost every project there comes a point where you need to cleanse your existing data to make it work with a new system. It is simply an element of how these projects work. You install or update a software package, attempt to fit your data into a new data model or attribute structure, and quickly find out that your data doesn’t fit the new system.
Truth #1 – Data Remediation is Rarely thought of as part of the Project Plan
This is the first mistake of data remediation: Not having a plan on how to deal with data is a huge mistake in any project. Yet many companies TrailBreakers works with have no data remediation plan. Either they assume the system they are implementing will fix the problem for them or assume the data will magically fit the new model. This is rarely the case.
We worked with a major durable goods manufacturer that wanted to improve its online sales by adjusting its product data model to be both more globally focused as well as removing redundancy in the collection process. There was significant remediation required because the current data model wasn’t collecting the needed information to go global. This required us to plan for the acquisition, normalization, and input of data into the system BEFORE the project started, or the entire project would be delayed as we attempted to fix the data. The entire project would have been delayed months without a plan. The plan for data remediation is the plan to keep the rest of the project on schedule.
The first lesson in data remediation is that you need to be planning for data remediation as part of the planning of the project. Waiting for data to fail to meet requirements is a huge point of failure your project can easily avoid. Knowing where your data is and the level of quality in that data is as vital as the rest of your project plan. At TrailBreakers, every one of our project plans includes a data remediation plan.
Truth #2 – Leaving Data Remediation Until After the Project Means Data Will Never be Completed
When data isn’t appropriately planned for, the next reaction is to leave remediation until after the project. This is the second mistake, as data that is not remediated during the project rarely gets remediated after the project. Loading the data “as is” because of a lack of planning for data will result in a decrease in the effectiveness of the project.
Leaving data remediation until after the project will also fail at the outset. TrailBreakers working with a plumbing supply company that wanted to install a new PIM platform. Their plan was to let the data that existed in their on-prem PIM work in the new tool. TrailBreakers demonstrated how advancements in the functionality of PIM utilities, especially around data quality, resulted in a lack of fit from the old data model to the new data model. Simply put, they could not simply shove the existing data into the new tool. TrailBreakers avoided a project disaster by anticipating the data changes needed and encouraging the customer to plan for that necessary activity.
Some tools will allow you to place data that does not conform to the data standards in the system. Although this methodology speeds up projects by allowing you to avoid the data remediation step, they create many other issues. TrailBreakers has never been part of a project where data remediation is budgeted after the project. Simply put, data remediation will not be completed if it’s not planned for.
Therefore, the second lesson of data remediation is do the data remediation up-front. Leaving it until after the project will limit the return on investment in the project, as improvements in data will not occur that are at the heart of the project’s mission. If you are planning on staking your reputation on the effectiveness of a project, take a stand on data remediation early in the project planning.
Truth #3 – Remediating Data More than Once is an ROI-Negative Prospect
The third mistake that people make in data projects is trying to remediate data over time. Remediating data once allows you to increase its overall value, creating sales lift, fewer returns, and a bigger overall basket value. Planning for data remediation during a project allows you to maximize your project’s return.
However, many times companies today take a point solution view to data modeling. As each new platform is added to the ecosystem a new data model is created, a data remediation (hopefully) occurs, and the cycle repeats. This point system view allows for quicker implementations but requires data remediation to occur multiple times. Each incremental remediation increases the data quality but comes at the cost of an additional review and cleanse of the data. Additional data remediations at best get you to the same level of data quality as a single well-thought-out project, but at a much higher cost.
The solution is an ecosystem data model view. Understanding this overall data model isn’t about implementing the exact same data model across every system: It is about understanding the data goals across that data domain and ecosystem so that every system in line with that data domain is adding to that overall data quality in a coordinated fashion. The data from a single system shouldn’t require remediation to fit the next system’s data requirements: The system should conform to the data requirements of the overall ecosystem data view.
The third lesson in data remediation is that you can maximize your return on your project investment by investing in an ecosystem view of data rather than a point system view. Understanding a broad view of data based on available sources and output requirements across the entire data journey is invaluable to the ease of transport of data, the maximization of your data as an asset, and your overall platform success.
In Summary
The TrailBreakers methodology prioritizes data as an asset, not an expense. Treating data as an asset prioritizes the data over the systems that source, transport, and display that data. Therefore, our projects look at the ecosystem to maximize value, not the system to minimize implementation costs.
These three truths may seem self-evident, but they come from many lessons during projects where companies prioritized cost over value. These projects taught us to look at the bigger picture of data and then apply the correct people, technology, and processes to maximize value. Our process is valuable because it leads to sustainable growth, not lowest-budget projects.
Therefore, a TrailBreakers project will look at the data holistically to avoid incremental data remediation. Every project includes a plan to understand data needs, sources, and curation to avoid excess data remediation cost. Our belief is that doing the project completely involves doing the project correctly, not at the highest budget, lowest budget, or fastest timeline. We are a partner and an asset in your projects: Not an expense.