Whitepaper: Data Onboarding

How to Avoid the Traps of “Dirty Data”


Savvy marketers know the value of matching existing first-party data such as CRM data, website registrations, or point-of-sale information to online data: more accurate targeting, increased campaign reach, and a better customer experience. But would it surprise you to learn that anywhere between 50-70% of the information you get from data onboarding is “dirty data?” These incorrect linkages and inaccurate matches generate tremendous waste in your media buys.

In this white paper, we’ll show you the pitfalls of many traditional data onboarding methods and explain how you can avoid them.

Trap #1: The Majority of Linkages Are Hypothetical “Filler”

Many onboarding providers start by directly connecting your offline data to online cookie pools—typically via your customers’ email addresses—without first corroborating your data with an authoritative offline dataset. Because they don’t have any independent data available to use to “clean” your records, they can only rely on the data you already have, such as your CRM. This often can contain incomplete and outdated information. As a result, they will find fewer of the possible accurate connections to online cookies or mobile advertising IDs.

Definitive email-to-cookie linkages typically generate a 30-50% match rate1 — not a very compelling business case. So, onboarding providers will try to extend their reach by supplementing your data with statistical modeling (which is really another way of saying “best guesses”). Naturally, these hypothetical linkages result in a higher rate of false positives, meaning that even if you’re told that you have a high match rate, 50–70% of your targeting may really be directed at someone other than the person whom you intended.

The solution:

Partner with an onboarding provider that uses an authoritative identity methodology to ground your data onboarding before trying to match it to cookie pools or mobile advertising IDs. This critical step:

  • Makes it practical to utilize any identifiers in your CRM, not just email addresses
  • Increases match rates by being able to take into account past identifiers that are no longer valid but still connect to a person in your CRM, not just the identifiers that are in use today
  • Reduces the need for hypothetical linkages and their resulting false positives

Trap #2: Individual Targeting Alone Leaves More Gaps

A number of onboarding providers tout “individual targeting” as a more precise form of data onboarding. Individual targeting certainly sounds more accurate than household targeting at first glance. But once you dig in, the methodology behind individual targeting quickly falls apart.

The reason is simple: actual linkages, such as that between your CRM data and cookies/ mobile advertising IDs are more difficult to make at the individual level than at the household level.2 Because onboarding providers make fewer individual linkages, your potential reach is lower. And to compensate, a provider that can only target at an individual level will try to beef up its match rates through probabilistic statistical modeling. The consequence is even more false positives and ultimately more media waste.

So the question is, would you rather spend your media dollars on less reach that has less chance of properly targeting the right person or on greater reach that almost certainly will connect to the right household? Unless individual gender or age targeting plays a primary role in your strategy, you are almost always better off targeting at the household level.

Trap #3: Omnichannel Measurement is Impossible Without a People-Based View

Because most onboarding partners are unable to provide a persistent identifier to their ad-serving partners, they support any media attribution (including closed-loop analysis or multi-touch attribution) without a significant amount of assistance from your other partners. Furthermore, we all know that consumers are omnichannel, using multiple devices and jumping back and forth between the online and offline worlds as they progress through their journey with your brand. If you haven’t integrated an authoritative identity methodology into your onboarding initiative, you lose any ability to recognize your customer across different channels. Without that, you can never properly attribute revenue or measure the customer journey.

The solution:

An onboarding partner with an authoritative identity methodology incorporated into the onboarding process will give you a true omnichannel view of your customers you need for an effective media mix strategy.

Trap #4: Onboarding is Not the End Goal but Just the Beginning

Most onboarding providers today do only one thing: onboard your data. But to actually make use of your data once it’s onboarded, these providers require you to bring in additional partners.

Aside from the logistical headaches involved, activation with additional partners almost always results in some amount of data loss; in other words, the number of available matches steadily decreases as you progress from one partner’s solution to another. Not only are your campaigns less effective, but also you won’t be able to tie your media exposure or frequency back to a unique user because your vendors lack a common identifier. So how valuable is your onboarding initiative if you reach only a fraction of your customers and can’t understand how they are interacting with your different campaigns?

The solution:

Choose an onboarding provider that also offers a data management platform (DMP) and measurement capabilities. This holistic solution allows you to segment your customers, activate target audiences across multiple channels and devices, and measure across all channels and touchpoints. With a single provider performing all of these critical functions, you reduce data loss and increase the integrity and usefulness of your onboarded data.

Seven Questions to Ask Your Potential Data Onboarding Partner

1. What is your anticipated match rate?

Understand what the match rate is and how your potential provider will achieve that rate. If the provider matches directly to online cookie pools using a limited set of fractional identifiers that you provide, it typically will have lower match rates and lower data quality. The provider may require an additional data onboarding partner to incrementally match off of other fractional identifiers.

2. Before matching to online cookies, do you first corroborate and/ or supplement my CRM data with authoritative offline data?

If the provider omits this critical step, you will achieve less reach and see less accuracy from your onboarding initiative.

3. Do you use statistical modeling to improve your reach?

If the answer is yes, expect to see many more false positives and greater media waste.

4. How do you accurately resolve various online and offline data elements (phone numbers, email addresses, and devices) to a unique identity?

If your provider doesn’t have a robust methodology for resolving multiple permutations of the same person, you will experience unintended overlap and lose your ability to properly retarget or sequence your online marketing.

5. Can you give your ad-serving partners a persistent customer identifier to facilitate closed loop and multi-touch attribution analysis?

If not, you will lose the ability to optimize your online media mix and to understand how your digital marketing is influencing offline customer activity.

6. Can your onboarding partner activate your data?

If not, you’ll need to find another way to activate, segment, and syndicate targeted audiences from the data you onboard. And your approach may dilute your onboarding ROI due to data loss.

7. Do you provide channel measurement, both online and offline?

If not, you will find it tough to understand how effective your audience targeting is or to measure success across both online and offline channels.

Take a New Approach to Data Onboarding

Neustar Onboarding utilizes Neustar’s authoritative identity methodology to increase both the accuracy and reach of your onboarding initiatives. At the beginning of the onboarding process, Neustar connects your data to our continuously refreshed dataset of more than 220 million adults and 120 million households and incorporates a persistent identifier to resolve multiple identifiers into a single person to enable accurate segmentation, sequencing, and measurement. Because this approach can make use of multiple identifiers in your data—both past and present—you get more reach, higher precision, and the most privacy-friendly data onboarding capabilities on the market. Neustar Onboarding connects to an extensive partner ecosystem that greatly reduces the low-quality linkages that lead to media waste.

With Neustar, onboarding is only the beginning. Neustar lets you segment, activate, and measure results based on that data—without the data loss (and headache!) that comes from multiple partners.

View PDF

Related Resources