Many have billed the CDP as the marketer’s new secret weapon. That may be the case, but in our experience, it’s often the analyst that benefits the most from the introduction of a CDP.
It’s the analyst’s job to understand changes in customer behavior. It’s the analyst’s job to use data to map the customer journey. It’s the analyst’s job to build algorithms that determine product affinity, estimate customer lifetime value, and predict churn. All of these require clean, usable customer data that’s been collected and unified from numerous sources. And this is precisely what a best-in-class CDP provides.
Analyst’s Day-In-the-Life without a CDP
By some reports, analysts and data scientists spend up to 80% of their timing prepping and cleaning data. 80 percent. That’s 4 out of 5 work days spent getting ready to do the work that will move their businesses forward. Much of the time is spent developing data transforms and SQL joins just to get access to all of the data they need to comprehensively understand their customers.
Unfortunately, this process doesn’t get easier over time. Important data lives in silos across the organization, and getting access to it often means waiting weeks, months, or even years for IT to prioritize an integration. At best, each new use case, campaign, or channel requires adjusting an already brittle ruleset governing customer identity. At worst, analysts from different parts of the business define their customers in wholly different ways, using only the data that’s easily connected.
The impact of siloed data ripples throughout the business, leading to time consuming, fragmented analyses whose results are difficult to integrate to decisioning layers and marketing solutions. And because the customer behavior and transactions are disconnected and spread across duplicate profiles, the models that power hyper-relevant personalization are less effective or just wrong.
We’ve seen the “garbage in, garbage out” story play out time and again. For example, we worked with a retailer that had been using internal business rules and fuzzy matching to unify their customer data manually. After Amperity stitched their data and resolved their customer identities, analysts discovered that a campaign built using their old data had been sending high value customers both active and lapsed marketing messages. Using Amperity’s CDP, the brand was able to map customers to the appropriate segments so individuals received the right campaigns.
Analyst’s Day-In-the-Life with a CDP
CDPs are designed to solve your customer data problems. This means they continually ingest data from any source, even those that are otherwise inaccessible. Then they resolve identities and unify data into rich, complete customer profiles – correctly associating customer behavior in the process. Many CDPs include pre-built attributes or models for common insights that enable analyses to be more repeatable. Finally, they give you direct access to explore that unified data directly, or send it out to your preferred analytics environments.
They also close the loop between analytics and marketing, allowing teams to operate in lockstep. Marketers know that they’re using the exact segments, attributes, and thresholds recommended to them by analytics because they have access to the same data. And the analysts are able to see the results of targeted segments without trying to reproduce a complicated campaign ruleset.
Remember that 80% of time spent preparing data? CDPs allow analysts to reallocate that time to higher-order thinking and insights. In short, a CDP should handle the boring and tedious data prep, so analysts can do the fun stuff that they were trained for: develop accurate, holistic insights that make conversations about the customer more nuanced, influence better decision making, and allow businesses to operate in a truly data-driven, customer-centric way.
The impact that an effective CDP can have on a business is monumental. We partnered with a retail brand to unify their online and offline data, making cross-channel customer insights accessible for the first time. Their analysts discovered that their fragmented data had led them to systematically undervalue their best customers. Now, with access to unified data, their analysts have generated insights that disprove long-held assumptions about their customers and identify new ways to drive high-value behavior.
Choosing an Analyst-Friendly CDP
Not all CDPs are created equal. Some trap data inside the platform, making it impossible to use your preferred tools and environments to run models. Others are so focused on enabling marketers to orchestrate campaigns that they lack the holistic data collection and resolution needed to truly know who they’re marketing to. Still others claim to unlock a single view of the customer, but only do so for certain channels, leaving analysts to pick up the pieces for the rest of your customer data – or not. Here’s a checklist to consider when you’re vetting a CDP that’s analyst-friendly:
Excels at customer identity resolution. Look for a CDP with advanced AI-driven matching capabilities, not deterministic, rules-based, and fuzzy matching. Traditional approaches are fragile and don’t resolve transitive identities (or work across sources that lack unique identifiers), which means they don’t deduplicate the most fractured identities.
Puts the data back in your hands. An effective CDP allows analysts and marketers to use the same data, removing friction between the recommendations analysts make and the actions marketers take.
Connects all of your data. For accurate customer insights, you’ll need data from all your online and offline systems – POS, eCommerce, loyalty, surveys, clickstream, email, etc. This works best when your CDP ingests raw data in its native format, every day.
Gives you the freedom to use your preferred analytics environment. Analyst-friendly CDPs give users the freedom and flexibility to both query and segment data inside the platform, and also to send data securely to external systems like Tableau, Power BI, and other best-of-breed analytics tools.
Gets smarter over time, and remains flexible. Let’s face it. The systems you love today might be outdated next year. A CDP should be designed for change, so as new sources of data and new marketing execution and analytics tools come online, you’re able to rapidly integrate them. A best-in-class CDP will learn from your data regardless of source and build richer, more accurate customer profiles over time.
To learn more about Amperity’s Customer Data and Identity Platform and how it’s empowering analysts at the world’s most loved brands to transform their businesses, visit amperity.com.