Whether your data is right or wrong, it's driving every decision you make.
Read that again.
Whether your data is right or wrong, it's driving every decision you make.
Let's face it. Maintaining quality data is a never-ending, thorny challenge that continually gets postponed or ignored altogether. Why. Is. It. So. Hard?!
It can feel like decluttering your garage. Where do you even begin? What do you remove? What do you keep? When do you get help? It's a ginormous project with no definite beginning or end. You know it needs to get better, but not sure how to rein it in.
Here are five common data quality obstacles and what you can do to ease the pain.
You don't know where to beginThe most common challenge we hear from our clients is that they don't even know where to start. We've heard this response from multiple clients:
"This is a really big project that we're not prepared to take on right now".It's only going to get worse. Putting it off will make it more complicated. The Solution: Identify and prioritize your data quality issues. What are your company's biggest pain points? Here are some examples of challenges we've discovered where data is secretly impacting results
Missed deadlines - Is it a personnel challenge or is the team struggling with data issues?
Poor marketing results - Bad content or missing/inaccurate/duplicate data?
Profit challenges - Blame sales? Or is it incorrect buyer profiles/customer persona?In order to prioritize, you need to identify where you're feeling the most pain. Maybe it's not that insidious, maybe it's as obvious as customers complaining about duplicate messages. Once you've identified the biggest pain point, go below the surface and do the deep dive. Ask questions and analyze the data for specific projects. "Sunlight is the best disinfectant." It's easy to assume the issues lie on the surface where we can see them, but it's critical to peel back the layers and get to the real causes. Using the above scenarios of common challenges, here are some examples of ways to do the "deeper dive". Missed deadlines - Step through and map each process in the data flow from beginning to end and pinpoint:
The input data
The output data
What is combined
What are the layouts
What steps are taking the most time
What steps need to be repeated because they were done incorrectly the first time (and second and third!)
Any QC steps that are missingPoor marketing results - Manually sort and examine the customer data that was used for the campaign and look for:
Outdated addresses, phones, emails, demographics etc...
Missing information (i.e. FIRST NAME field populated with "Unknown")
Duplicates
Processes that removed "good" records and keep the "bad" records
Wrong Information
Information in wrong fields (i.e. LAST NAME is in the FIRST NAME column)
Data Integrity issues (columns skewed by commas or human error)Profit challenges - Identify and review the data used for customer personas (you are creating customer personas, right?):
Pricing - if you're customer data has quality issues, how do you know if your products are priced correctly?
Customer acquisition cost - Multiple vendors and processes will increase your acquisition cost leading to decreased profitability
Inability to accurately determine the customer lifetime value
Your data Is not standardizedThis may sound basic but plagues most customer files.
Multiple fields with the same info (Multiple phone fields, multiple emails)
Different data sources that need to be combined with fields in a different order
Input standards not enforced (may not be able to be enforced)
Data in the wrong field
Different teams with different standardsThe Solution: Even though the input data for multiple sources will have different layouts, it's important to agree on basic standards. During any processing, data that does not meet standards are updated to comply. Also, to ensure standards are always met, the process must be automated. Any vendor you use should have the ability to fit within your required standards.
Your team currently runs multiple, ad-hoc, manual processesDespite having the right tools, teams will implement whatever process they need to get their work done. For example, one week a team may dedupe before running NCOA and the next week, run NCOA first before deduping, producing unexpected and unreliable results. The Solution: Set up automated processes. Start with one process to avoid being overwhelmed. Your existing tools may allow for automating the needed steps or it may require custom programming. Oftentimes, IT teams have that capability but may not have the time. If that is the case your data quality vendor should be able to help you.
You use multiple vendors for various services A reactive approach to data quality consists of disparate data hygiene steps: Choosing different vendors for each process, sending your data to each vendor, receiving back different layouts, having to further manipulate that data, and conducting some additional cleanup in-house. What are your answers to the following questions?
What is the minimum cost for each of your data processing vendors?
What was the total cost of your most recent list processing?
Where do the bottlenecks occur in your data preparation process?Using multiple vendors for various services is tip-toeing through a minefield. The Solution: Use one vendor to handle the data quality end to end. Choose a vendor that takes into account:
How the data is collected
What type of data it is
is it from a sale?
Is it from a customer or prospect?
Are employees entering information?
What is the goal of the data?
What are the planned uses?
What are the real issues?Instead of using different, laborious processes for solving data problems, these solutions are customized to fix quality issues and ensure ongoing data accuracy and consistency. Custom solutions are based on data sources, industry priorities, and individual company goals, and their interconnectivity allows you to unify information across different teams and divisions of your business to ensure decision-makers at every level are working from the same — accurate and reliable — data. Data quality solutions simplify day-to-day processing and the business operations that incorporate data.