Data hygiene may not be the most exciting topic in marketing, but its value is immediately apparent as soon as any marketer tries to use unreliable data. Omnichannel marketing, marketing attribution, media spend optimization or any of the other marketing strategies and goals required to succeed today all require accurate data.
If the data isn’t accurate, optimizations, reporting and attribution records won’t be accurate either. This can result in personalization efforts that alienate customers rather than engage. If the problem is widespread enough, expensive marketing programs may fail, not because the strategy or implementation was wrong, but simply because enough of the underlying data was inaccurate. Ultimately, this can lead to loss of revenue as marketers are unable to accurately optimize campaign spend for successful or underperforming efforts.
As Rex Briggs, Founder and CEO of Marketing Evolution, explained recently,
“All the data in the world is useless if marketers are measuring the wrong things, and may cost marketers billions of dollars in wasted resources.”
Recent research reveals that data itself is often the key barrier to implementing a cutting-edge marketing strategy. Whether the issue is difficulty integrating, managing or collecting it, data is often both the best asset and the biggest challenge modern marketers face.
Fortunately, achieving better data hygiene is an attainable goal that produces considerable rewards.
How Marketing Data Quality Becomes Compromised
Sophisticated marketing systems often produce sophisticated problems. Some of the most common data issues marketers run into are:
Out of Date
Data can become inaccurate alarmingly fast. For example: 17 percent of Americans create a new email address every six months.
One data input system may record customer website visits one way, while another input source records them in a different way. Without intermediary software to make the formats consistent, the data isn’t usable.
Overwritten at the Wrong Time
Whether it’s human error or machine error, if even one field gets skipped in a software routine, the entire data output can be wrong.
Duplicate records are the bane of most sales departments, where one prospective customer contacting the company multiple times can result in multiple files in the lead generation database.
For whatever reason, some data just comes into the system wrong in the first place. The classic example of this is of misspelled names. Brands can spend hundreds of dollars trying to convert a high-value prospect, only to realize that every communication they’ve sent has been dismissed outright because the prospect’s name wasn’t spelled correctly.
Information fields can be left unfilled or become blank for myriad reasons. Sometimes this renders all other information for that file useless.
If you are working with legacy systems, or have inherited data files from someone who left the company years ago, it’s quite possible to find fields and codes that don’t make sense. Or, information in the fields may not really convey any truly useful information.
This is one of the most difficult data hygiene problems to solve. Assessing which data fields are actually being used by different teams and individuals inside or outside of the marketing department can be real detective work. But it’s usually worth it: unused data fields bloat data files and slow down optimization. If certain information isn’t being used, and isn’t going to be used, it should be deleted. Due to GDPR, this is often actually required by law.
The Role of a Data Audit in Data Hygiene
The first step to overcome these challenges and improve data quality involves a data audit to reveal the scope of any problems. The data audit may find all of these problems, or just a few. It may find that most of the data in a file is accurate, but that certain fields need to be re-verified.
Kristin Hambelton shares some helpful questions for marketers to ask as they begin a data audit in her blog post, “How to Trust Your Data Again”:
• “Where did this data come from?”
• “Why does this data matter?”
• “How will it be used?”
• “How will it drive value and increase ROI for this campaign?”
• “How will it be maintained and validated for accuracy?”
A data audit’s ultimate goal is to answer questions like these. It should aim to deliver the person-level data view required for optimal results.
Data audits often involve other departments connected to marketing, including sales, customer success, customer service, finance and the C-suite. This usually requires robust coordination between key contacts in each department, and may trigger smaller data audits within those departments.
In some cases, it may also make sense to purge old data fields. Computing capability is not unlimited, and if a marketer has to pay to keep each file updated, the fewer files, the less the data hygiene maintenance will cost.
Once the scope of data corruption is understood, and the existing and future data needs of the department and the broader company are mapped, there is often a need to update or fill in several data fields. This can be done by directly recapturing lost or inaccurate data, or it can be done via third-party services with data overlays. A blend of the two approaches usually works best.
How to Maintain Marketing Data Hygiene Going Forward
The work isn’t done once a database or data system has been cleaned up. Now the job is to maintain good data hygiene. This requires multiple systems and best practices, including:
Gathering data from the right sources
Leveraging the right marketing analytics platform to separate, manipulate and process new and existing data
Developing the right measurement models to interpret data
Establishing the appropriate optimization processes to leverage data for optimization
Today’s technological capabilities would astound marketers from even twenty years ago. And while unified marketing measurement and optimization can produce impressive results, it must have accurate data to work.
Without accurate data, expensive campaigns fall short. Get your data right, and today’s marketing initiatives will start to work considerably better.