Data services companies like Acxiom, Merkle, Experian, and Epsilon play a critical role in helping analytics and marketers understand their customers – by enriching profiles with consumer data. But are they the right choice for building and maintaining your brand's single customer view? To find out, start with an honest evaluation of your current customer data foundation. This will help you determine if it’s capable of supporting all your customer-centric goals. Here are the top three questions to ask yourself and your team:
Are all of the data sources our teams need for customer-centric insights and initiatives centralized, unified, and usable?
Is our customer database accurately deduplicated at the individual level?
Is our data ready for analytics, machine learning, and syndication to marketing tools?
Before we get into addressing these questions, though, it’ll be helpful to give a quick overview of how data services providers and CDPs differ when it comes to customer data management. You can also download our quick guide CDP vs. Data Services Provider: a step__-by-step guide to understanding your options for customer data management to learn more.
Data services providers are specialists in procuring, maintaining, and leasing or selling consumer data. Some also provide marketing advisory and services. Many have been around for decades. Over the years, they have expanded their offerings to include some identity resolution and customer data management, mostly as custom-built solutions. Data services providers added these to help brands get more value from the 3rd party data they were selling.
A Customer Data Platform, on the other hand, is designed exclusively for the data management challenges associated with today’s 1st party customer data. True CDPs are software-as-a-service (SaaS) products built to handle the scale, speed, and complexities of customer data, so enterprise brands can improve insights, personalization, and omni-channel measurement. They rose to popularity because nothing currently available was working.
And now, we return to...
The Top 3 Questions to See if a CDP is the Answer
1. Are all of the data sources our teams need for customer-centric insights and initiatives centralized, unified, and usable?
Data services providers’ solutions were designed for a different era, when data was simpler, smaller, and changed less frequently. As a result, they typically build static databases with fixed schemas that quickly become outdated as new data sources or systems are introduced. When teams try to incorporate these new sources, they face lengthy waiting periods and re-builds, or worse, contract renegotiations and up-selling. These databases also struggle to handle the scale of modern big data, forcing users to wait hours for queries to run or preventing concurrent use.
Most CDPs, in contrast, are cloud-native platforms, architected from the ground up to take advantage of cloud efficiencies, flexibility, and scale.
For example, one Fortune 500 retailer we work with had email response data and online transactions available for analysis and other use cases in the customer database their data services provider had built. But they were missing loyalty data, call center interactions, and digital clickstream data. Adding these new sources would have been a slow, expensive process with their data services provider, leading them to invest in a CDP instead.
2. Is our customer database accurately deduplicated at the individual level?
Data services providers often offer identity resolution alongside data appends for sparse data. This can be useful. But when it comes to building a 1st party customer identity graph, they rely on older approaches like rules-based deterministic matching and fuzzy matching. Over time, these approaches degrade as the complexities of real-world data stack up, resulting in a customer database riddled with over- and under-resolved records.
In one study we saw that brands were, on average, under-clustering 22% of their customers and over-clustering 1%. Worse still, these were many of these brands’ best customers, responsible for over half their total revenue. These inaccuracies mean even the most basic customer insights (like who are my best customers, how many customers do I have, etc) are wrong, and so are targeting, personalization, and cross-channel campaign measurement.
One Amperity customer, after working with a data services provider for several years, found that their database has devolved such that about 50% of their customer profiles were inaccurate (due to a combination of over- and under-resolution). An advanced identity resolution solution was required to keep their data foundation healthy and durable: one that uses artificial intelligence to cluster records together and keep identities stable over time. Some CDPs specialize in this area and can provide significant improvements in identity accuracy and stability over a data services provider like those outlined above.
3. Is our data ready for analytics, machine learning, and syndication to marketing tools?
Every use case and system has different requirements for how data is merged, shaped, and syndicated. Because of the static nature of the extract-transform-load integrations and databases that a data services provider builds, it's difficult to quickly and cost-effectively update schemas and custom-built connections to source systems in order to accommodate new ways of using data.
CDPs are agile, SaaS products that are constantly being iterated on, added to, and upgraded with every release. Typically it’s faster and easier to make changes to the shape and formatting of data for new use cases and out-of-the-box integrations with new channels and systems are continuously being added.
For example, one Fortune 1000 brand wanted an easy way to land their unified customer data in Tableau for analysis, but their data services provider hadn’t formatted the data correctly and didn’t integrate with that platform. Adding additional destinations would be slow and costly. They turned to a CDP with an existing integration with this common tool.
If any of the questions above raised concerns about your customer data management capabilities, it’s probably time to start evaluating CDPs or other customer data management options.
Playing Well Together
Even if your data services provider doesn’t solve all your customer data management problems, they are still likely a valuable partner. For many businesses, their 1st party customer data has gaps that can be corrected with the 3rd party data.
The best way to couple the use of 3rd party data with a CDP is to first organize, resolve, and assess all of your first party data with the CDP. Then, as you identify actual holes in the data, partner with a data services provider to fill them. By first properly resolving identities, customers’ most accurate and up-to-date data can be used to match against 3rd party data sources, improving match rates. CDPs can ingest and combine 3rd party data with 1st party data for even more holistic customer profiles.
This approach means spending less on data overall (because when it’s finally all unified you often have a lot more information about your customers than you thought), and leaving customer data management to the true experts – CDPs.