The Clean Data Challenge
- Rolf Olsen
- Jul 16, 2021
- 2 min read
Updated: Jul 6, 2022

Over the past couple of weeks, I have seen numerous posts talking about the challenges facing analysts and data scientists regarding dirty data, and specifically who should clean it and why.
Having worked on both reporting and modeling projects over the years, I can safely say that I understand the pain! Bad data does not lead to any kind of useful report, insight or model, period.
The reality here, is that the task of “fixing” the data often falls on the analyst, who has to scramble to try to generate an output for deeper analysis.
While there are numerous use-cases for various types of businesses and verticals, I am going to focus here on the world of media and marketing data.
Cleaning the data is certainly an option, when possible, but it is laborious to say the least, and importantly, it cannot be done in isolation. By this I mean that the analyst or data team cannot simply look at the data and fix it. They are at the end of the line as it relates to the data, which makes it hard to correctly categorize and organize all the available input, which is particularly important when looking at a multi-channel media strategy.
While multiple investment and trafficking teams are normally involved in generating the data, these efforts are often not well coordinated and comprehensive data strategy is not fully taken into consideration in order to enable the correct mapping for a cohesive merge of all the data.
This brings me to the the followings questions... What if we did not have to clean the data? What if we had a foundational data operations and governance team, that could guide the underlying data strategy and provide the necessary tools and processes to minimize errors by easily identifying potential issues and connecting with the appropriate teams to ensure fast and systematic data correction?
Notice how I did not mentioned analysts? They would work with the data operations team to ensure their data needs are clear for the specific tasks the data is needed, rather than blindly trying to jam square pegs in round holes.
So, if you are looking to create timely and accurate reporting, and turn around faster modeling projects and more reliable insight for better planning, optimization, and forecasting, then the solution is clear... Build a data operations and governance team!
"Data Governance is a system of decision rights and accountabilities for information-related processes, executed according to agreed-upon models which describe who can take what actions with what information, and when, under what circumstances, using what methods."
– The Data Governance Institute
If this is something your company needs, give us a call or send us a message and we will work with you to end the reign of bad data and help your analysts do what they do best, empower your organization with meaningful and reliable reporting and analytics.
#marketingdata #datagovernance #dataoperations
Comments