The Crises and Cost of Bad Data

More Is Not Always Better

The rise of data-driven marketing has led to a “more is better” mentality to inform better and better marketing. Why? Brands seek out ever-growing data sets to reach more customers, make content or messaging more relevant, help campaigns perform better, solve complex problems, and help deliver better customer experiences. Some use data to help sell more advertising inventory, or their purpose in seeking more data is actually to sell the data. Most also seek solutions for cross-channel data measurement and attribution.

It’s becoming apparent that more is not always better where collecting data is concerned. The benefit of such data does not always manifest, and the data landscape is changing rapidly. Some data sets we’ve come to rely on have changed or are no longer reliable or available, and the risk and true cost of such data have become increasingly apparent. Several key issues drive the crises of data:

• Looming Government Regulations
• Privacy and Brand Equity Risks
• Changes and Issues with Third-Party Data
• Backlash Regarding Data Sellers and Sharers
• The High Cost of Bad Data

Looming Government Regulations

Increasingly, consumers are paying more attention to their data footprint and their data rights. They’re realizing it’s a valuable resource and are starting to ask more questions about how it’s collected, who has access to it, and how it’s used. Inevitably and unsurprisingly, governments are not only starting to pay more attention but also acting on what they perceive to be industry issues and abuses.

The European Union’s General Data Protection Regulation (GDPR) was the first real wake-up call in terms of overarching government regulations. It is the most substantial change in data privacy regulation in decades and has led to extensive modifications in data collection and consumer-facing messaging. While consumers may not understand what GDPR is, they certainly have seen the “this website uses cookies” pop up, which ensures compliance.

GDPR has far-reaching impact beyond the borders of the European Union and is just the tip of the iceberg. The Brazil General Data Protection Law (LGDP), the California Consumer Privacy Act (CCPA) set to go into effect in 2020, and multiple other states contemplating their own privacy legislation make data collection and usage more complex and, dare we say, riskier than ever before. This leaves marketers concerned about how they will derive value from their existing first-party data investment and how they should (or even can) continue to use third-party data given a patchwork of stricter regulations.

The possibility of stronger government regulation is perceived as the main potential threat to “deriving value from data,” according to 53% of industry executives (IAB).

Privacy and Brand Equity Risk

GDPR has done significant good in raising awareness around individuals’ rights. This means every single company is risking its brand equity and reputation if it doesn’t handle data appropriately. Smart data collection and usage policies certainly must be put in place to avoid running afoul of regulations, but they are also necessary to ensure marketing does not cross the line from personalized to creepy or abusive. Twenty-five percent of consumers have stopped doing business with a company specifically due to poorly personalized communications. This number grows to 35% among Millennials, according to a study by Broadridge Financial Solutions. Brands must think strategically about how – and how not – to use their data assets when marketing.

Protection of data assets is critical to prevent breaches of personal data, which can lead to an unrecoverable hit to brand reputation. An issue like this will linger long after fines are paid. For most organizations today, the question isn’t if there will be a data breach. The real question to be answered is when it will happen. Because a company’s brand equity is on the line, having a plan to respond quickly and appropriately is critical. Trust is more important than ever for brands. Knowing how to deal with customer data, especially when issues arise, is essential to engender trust.

Changes and Issues with Third-Party Data

Some say the accuracy of third-party data is overrated and we have passed the point of diminishing returns on further modeling. Data sourced from others can be faulty, inaccurate, out-of-date, or utterly irrelevant to a brand’s business goals.

Relying on third-party data purchasing alone is a suboptimal practice. The data can be amassed from a wide variety of independent and unreliable sources like credit scores, cookies, and click trails. It can quickly become outdated and antiquated as preferences, purchases, and household priorities change. In one ChoiceStream study, a particular data vendor had identified 84% of users as both male and female. While the case was an extreme outlier, another Mediasmith report on third-party data found that data from four of 11 vendors wasn’t much better than chance in targeting age and gender.

Are we being overly dramatic? Look yourself up on Oracle Data Cloud’s registry tool to see what audiences you belong to within their third-party data sets. They are one of the better players, but even with them, there is a high likelihood of contradictory or wrong data.

Why do these problems exist? Parties across the marketing ecosystem are to blame.

• Some marketers are guilty of “stuffing” their customer relationship management (CRM), marketing automation tools, and data management platforms (DMPs) with third-party data sets from aggregators (more discussion about them later) because it’s faster and easier than building out a robust first-party data strategy and program.
• Since scale is essential to marketers, data middlemen and brokers have a clear incentive to grow the size of their data sets. They are paid on volume, not on accuracy. Full transparency into the level of confidence and statistical significance of the data is not always what it should be. There are good and bad players in every industry. Marketers must invest the time to understand, evaluate, and question methodologies for collecting and refreshing data. Otherwise, it offers vendors incentive to misbehave.

Backlash Regarding Data Sellers and Sharers

To some extent, regulations and concern about privacy have come together to place into question those groups aggregating and selling data as a business model. While there are certainly good and bad players, the actions of the bad players dominate the discussion. Why? The media is prone to say, “Bad news is inherently better than good news.” It sells.

There’s no mistaking, however: there are a lot of bad practices with companies whose entire business relies on selling secondhand or third hand data. Data can be easily copied, disclosed, and fed to others without the consumer ever knowing. The data may have even been obtained without consent or an understanding of what data was being collected and why.

Even if data has been collected properly, usage of it can come under fire. A recent report from Motherboard revealed a person’s cell phone location could be pinpointed through data easily accessible from data resellers. The report led 15 US Senators to call for an investigation into how telecom firms handle their customers’ location data. The incident also led AT&T, Sprint, and T-Mobile to pledge they’ll stop selling location data to data aggregators.

Of course, no discussion on this topic could be complete without discussing the Facebook-Cambridge Analytica data hijacking scandal. In early 2018, it was revealed that Cambridge Analytica harvested personal data from close to 85MM Facebook profiles without consent then used the information for political advertising. Yikes! Facebook does not sell data but instead uses data to sell access to people. Data sharing and data selling may not seem too different, but they are. If Facebook were to sell personal data vs. sell access to people, they would be giving up a ton of value. It may be a fine line, however, as there remains a considerable amount of concern and issues with the model while Facebook continues to learn.

The High Cost of Bad Data

The Interactive Advertising Bureau (IAB) and Winterberry Group estimated that US marketers spent nearly $5BN on data management and integration in 2018. This includes investment in products such as CRM systems, DMPs, customer data platforms (CDPs), and identity resolutions. It’s further estimated that US companies spent more than $19BN on audience data acquisition and solutions to manage, process, and analyze the data.

As digital marketing becomes more automated and data-driven, it has become crucial for its practitioners to corral the data on which they rely. It’s essential to understand what information is beneficial and what should be tossed. Unfortunately, there’s no silver bullet for separating useful data from bad.

Bot attacks are on the rise. Companies must protect themselves and their users’ data from malicious attacks and bad data creeping into systems along with the good. The IAB estimates online ad fraud costs advertisers $8.2BN annually.

Then, there’s the cost if bad data enters an organization’s system. If the goal of data-driven marketing is to inform an organization’s efforts better, bad data can throw a wrench into the system, causing a whole host of issues:

• It can lead to wrong assumptions, wrong audience insights, and flawed initiatives.
• Effectiveness of marketing campaigns can be negatively impacted; in extreme cases, poor engagement and declining open rates for email can lead to your communications being classified as spam.
• While it may be just cents per transaction if bad records exist, the cost will add up throughout multiple campaigns.

It is important to ensure an organization’s pipeline of data is verified and validated BEFORE it makes its way to your data warehouse.

What does all this mean?

Data management and data-driven marketing are challenging to implement. They require a network of partners (internal and external) all working from a single playbook. To do this well, marketers – or their agency partners – need the right skills, experience, and methodology for evaluating partners. A strong ability to determine the best, highest quality sources of inventory and data is no longer a nice-to-have; it’s necessary. Further, these trends provide the impetus for marketers to create and nurture direct relationships with their consumers and customers rather than lean on third-party data obtained from others.