The latest ANA benchmark shows that only 43.9% of every $1,000 that enters a DSP reaches consumers. That’s progress compared to previous years, but it still means more than half of ad spend is lost before impressions are delivered. The issue comes down, in large part, to poor data quality. When advertisers tap into first-party data that’s inaccurate, incomplete, or stale—be it for campaign strategy, media buying, audience building, or targeting—they’re wasting media budget and tanking campaign performance.
During the dawn of the big-data era, data quality was considered a nice-to-have as focus was largely on how much data an organization could acquire and what could be done with it. However, we’ve entered a new chapter of that era—one in which brands cannot let data quality fall by the wayside. Today, CMOs increasingly demand that teams drive measurable business outcomes. Moreover, technologies like AI, powered by sophisticated algorithms trained on large datasets, are reshaping and redefining media buying as we know it.
Understanding data quality in advertising, why it’s essential to campaign performance, and how advertisers can enhance it through proper data hygiene practices can mean the difference between success and failure in today’s world. This helpful explainer will give you a data quality foundation, so you can:
So, how do you know if your data is up to snuff? There are six dimensions of data quality that brands should examine: accuracy, completeness, consistency, timeliness, validity, and uniqueness. To illustrate how these dimensions work together to indicate data quality, we’ll follow a hypothetical luxury coat brand, Tundra, in its journey to drive sales.
Consider the various elements of a customer database that a marketer simply cannot afford to get wrong, such as email addresses, home addresses, and device IDs. If your customer data is inaccurate, this can impact reach, reputation, media spend, and campaign performance.
Tundra aims to target customers in northern US states who haven’t purchased a coat in five years with ads for its newest line. The brand segments its customer list by shipping zip codes in the northern US. However, the brand never verified the accuracy of its data or updated its records prior to launch, and as a result, it mistakenly targets audiences who have since relocated to warm-weather climates. By not checking data accuracy prior to launch, Tundra misspent its media budget and potentially created a negative brand association among those it mistakenly reached.
Data completeness is a critical component to informing strategy, targeting, and delivering outcomes. Without complete data, brands may miss out on opportunities to engage consumers with relevant, personalized experiences—or at all. Let’s examine another example.
A woman likes to visit her local Tundra store on her way home from work, making occasional in-store purchases rather than shopping online. Because her in-store transactions aren’t linked to her online customer profile, her digital profile is incomplete. When Tundra sets out to re-engage lapsed customers with an email campaign offering a discount for the season’s newest styles, she receives a discount offer for coats and accessories she recently purchased. This incomplete, disconnected data leads to wasted outreach and a poor customer experience.
Consistency in formatting impacts audience segmentation and targeting, frequency management, measurement, personalization, and optimization.
Say Tundra launches a regional campaign to reach customers in Massachusetts. They segment their database to include only those with “MA” as their state of residence. However, without any data normalization practices in place, Tundra mistakenly excludes people who live in Massachusetts, even if their state is listed as “ma”, “mA”, or “Ma”. This significantly impacts Tundra’s reach and scale, limiting ROI and skewing campaign results.
While data completeness is an important dimension to examining one’s own first-party data, it’s also a critical component to understanding the quality of the media one buys. For example, Tundra may want to run ads on CTV. The organization taps into different inventory sources, but one CTV app lists inventory sourced from a network called Standard TV as “STND network”, while another lists it as “STND-ntwk.” This inconsistency in data inventory signals causes duplicative bidding, wasted media spend, and inaccurate reporting.
Data is only as useful as its recency. Whether working with data collaboration partners, data marketplaces, or one’s own first-party data, leveraging real-time signals can mean the difference between a highly successful campaign and a failure. Brands need access to the most up-to-date information about their customers to ensure they reach them with relevant ads that account for where they are on their journey and all consumer touchpoints. Let’s look at another example.
Tundra kicks off a holiday campaign to reach individuals and households with ads for its newest women’s line, advertising across DOOH, display, linear, and CTV in the NYC metropolitan area, encouraging online and in-store purchases. However, Tundra fails to reconcile in-store purchases with online and digital profiles in a timely manner. Unfortunately, this lag causes Tundra to waste media, reaching customers who’ve already purchased coats in-store with ads across linear, CTV, and mobile.
Data validity refers to whether information adheres to established formatting rules, syntax or classification standards needed for accurate activation and reporting. When data is invalid, it affects audience targeting and campaign performance, yielding misleading results.
Tundra is running another CTV campaign; this time, it will appear only in premium lifestyle and fashion content, aligning with the brand image. Once again, the brand taps into its preferred CTV app for inventory, but much of it is mislabeled as lifestyle content when it's actually comedy. This data is invalid and will reach incorrect audiences, wasting media spend. Moreover, if optimization models are trained on invalid data signals, lifestyle genres will be deprioritized due to underperformance, driving Tundra to stop purchasing media on lifestyle content altogether.
This example illustrates how data quality and validity can impact both short-term and long-term success, such as campaign delivery, optimization opportunities, and insights that can inform future strategies.
Data uniqueness ensures that records in a dataset appear only once, with no duplicates, and that each record is distinct.
Tundra builds an audience for its winter campaign using data from its website, email list, and loyalty program. Without identity resolution to unify those records, the same customer may appear multiple times under different identifiers—once as an email subscriber, once as a site visitor, and once as a past purchaser. These duplicates inflate reach metrics and cause the same person to receive the same ad repeatedly across channels. By resolving identities into a single, accurate profile, Tundra prevents wasted impressions, improves frequency control, and gains a more truthful view of campaign performance
Understanding the six dimensions of data quality is only half the equation, however. The next step is putting those principles into practice.
With a foundational understanding of data quality and its six dimensions, you’re now better equipped to make the most of your data in your advertising efforts. Below is a practical checklist to help you strengthen data quality within your organization and across your advertising campaigns.
As marketing and advertising grow more complex, technology now plays a critical role in maintaining data quality at scale. The right platforms help unify, standardize, and activate data seamlessly across channels—and do so responsibly.
These six dimensions of data quality underpin every activation strategy and, together, comprise a seventh, unofficial—but equally as important—dimension: trust. Data quality builds trust between brands and their data collaboration partners, publishers and advertisers, and perhaps most importantly, between brands and their customers. Each side of the advertising ecosystem has an obligation to uphold data quality and integrity, as every player contributes to the industry's overall success.
Cadent helps uphold that standard. With Cadent Audience Manager, brands can seamlessly onboard first-party data and, using the Cadent Viewer Graph, match it to identifiers from offline and online channels to an individual or household within minutes—all with privacy and compliance at the core. This process, also known as identity resolution, empowers advertisers to connect with consumers reliably across more than 125 million households, 510 million emails, 740 million home device IPs, and 1.8 million connected devices.
With these capabilities, brands can engage audiences meaningfully across channels with relevant, personalized experiences and measure outcomes with confidence. Moreover, the Cadent Platform implements data standardization and normalization across key channels and formats to improve data quality, thereby positively impacting media buying and measurement.
Want to learn more about how Cadent can enable you to derive greater value from your first-party data and media budgets with improved data quality? Reach out to us today to discuss.