How to Navigate the Perilous Path From Data Collection to Data Activation

Stetson Barnhouse

How to Navigate the Perilous Path From Data Collection to Data Activation

In my previous article, we took a deep dive into customer data platform (CDP) success factors, focusing on your MarTech stack. Now you’re ready to actually get work done with your data.

First, a quick reminder that a CDP does two things. It collects and unifies first-party customer data (enhanced by third-party data when available) to provide an accurate, uniform view of each of your customers. It then makes that data available to create personalized marketing campaigns and optimized digital experiences.

Before we get to the tips on navigating pitfalls, let’s take a look at each of those capabilities.

Aggregate Customer Data

The data you’ll collect includes data you “own,” and data from every other source – first-party and third-party data, respectively.

First-party data is observed or provided directly from the customer. This includes their behavior on your digital properties, their subscriptions and purchases, and information volunteered on-site or through a call center. Think of your first-party data as the batter of your CDP “layer cake.”

One layer of this foundational data is the behavioral data coming from a tool like Adobe Analytics or Google Analytics. Another layer is transactional data, like which folks who visited your site eventually made a purchase in a store.

Following the cake metaphor, you might guess that the third-party data you acquire and append to your customer records is the buttercream icing on that cake. This data, coming from sources you have contracted with, enhance your understanding of your customers and prospects. It may tell you what other sites your customers visit or the ads they’ve viewed. Vendors and partners are the sources of third-party data.

Once your CDP has collected the ingredients of the cake, it uses a process called customer data integration to stitch together all your data, at a customer or business-account level. Put simply, CDPs take data from various data sources and store it in a secure, central data warehouse. There, the data is standardized, de-duplicated, and made available for analysis and activation. Both analyzing the data and activating upon it requires specialized systems we’ll cover below.

Empower Marketers to Activate on that Data

A CDP is designed for marketing. That marketing occurs once the various sources have built a 360-degree view of each customer. This is where the true return on the investment of a CDP takes place. With it, marketers can create highly personalized and targeted marketing campaigns and user experiences.

CDPs also support customer data management by organizing customer data in a user-friendly and commonly agreed-upon – and universally understood – format. This last point, a universal understanding, is key, since delighting customers “takes a village” and there must be consensus and close collaboration among your teams on what you’re all seeing in your CDP data.

With the right customer analytics system in place, you can better recognize intent. This may require machine learning tools or advanced segmentation that only Big Data can reasonably deliver. With every wave of new analysis, your understanding of your customers deepens and your understanding of customer intent becomes more accurate.

Sensing the intent of your customers and prospects, you can then employ activation systems to turn that understanding into improved customer retention and business value. Activation systems can include those that communicate with users in periodic (weekly, for instance) emails or real-time triggered emails, SMS messages, or in-device notifications.

Beware of the Pitfalls!

Vague Role Delineations and Sub-Par Training

Any CDP can handle the aggregation bit, but most stop at the generation of insights beyond the very basic. Members of analytics teams are often expected to wear a number of different “hats” — developer, domain expert, product owner, and QA expert — in addition to fulfilling their analysis roles of reporting and analyzing. This can stretch a team thin, ultimately reducing CDP effectiveness and lowering the pace of discovering true, actionable insights.

What’s worse, the investments in formal certification and interdisciplinary training are frequently one of the first items on a budgetary chopping block (if they were ever in the budget in the first place!).

Remedy: Establish a clear delineation of roles and train staff appropriately.

Poorly Defined Processes

Seamless coordination with execution channels, and dedicated campaign orchestration processes and systems, are critical to the success of activation of any kind, and even more so when using CDP data that required a significant investment to set up and license. Typically, teams are expected to do insight generation and customer activations at a frantic pace to meet dynamic marketing conditions.

In our experience, the most successful CDP capabilities are those that have adopted clear, repeatable processes built into a workflow management tool that removes emails, Excel spreadsheets – and even Slack or MS Teams – from the CDP work equation wherever possible.

Remedy: Establish clear processes. Extra points if you manage the generation of insights and personalization using a workflow management tool such as Adobe Workfront.

Inability to Prioritize CDP Requests

No organization we’ve worked with has enough resources within its analytics capabilities to handle the flood of requests that will come in from all corners of the enterprise once the CDP is stood up. That means prioritizing the requests is essential.

Remedy: Create a mutually agreed-upon and aggressively enforced measurement “North Star” strategy.

Prioritization is easy if there is a yardstick against which a request is measured. That yardstick is Measurement Strategy! Here’s how it works: The role handling requests compares each new one against the Measurement Strategy. If the request doesn’t contribute to a strategic goal, it gets moved into the project backlog.

Start with the primary purpose of your business. It’s very important to establish clear objectives, as it will make all marketing, branding, and content creation efforts fall in line. Next, establish SMART goals. (“SMART” stands for specific, measurable, achievable, realistic, and timely). These goals should be periodically reviewed to track progress by the Marketing Data Governance Body (see below).

Flowing from these goals are the strategic reports required to make the business decisions necessary to move the needle. What sort of business decision? If a goal is “improve engagement with FAQ content by XX% year-over-year,” there may need to be a resource allocation decision to invest more in the content of those pages.

Lack of Governance

The work described in the prioritization efforts above cannot realistically happen unless it is empowered by a governance team. Many times, some of the least strategic activations and reports are requested by high-ranking folks within an organization (don’t ask us why – It’s just the truth!).

Remedy: A governance body – a group empowered to make final decisions on things like prioritization – can provide this clout.

Create a cross-disciplinary, multi–business unit committee that meets regularly to consume the latest insights generated by the CDP and make decisions on activation activities, A/B tests based on CDP-generated segments, and even hiring and platform licensing needs. Conduct the meeting as you would any other steering committee activity, including maintaining minutes and voting on the decisions that come up where consensus isn’t possible.

Losing Sight of How to Act Upon the Data Provided

It’s one thing to create the warm feeling of producing campaigns and personalizations and reviewing reporting about them. It’s a far tougher thing to make clear, data-inspired decisions based on that activity.

We believe a report or campaign finding isn’t an insight unless it is the following:

Statistically valid
Pointing to a change that would improve business outcomes
Acted up

OODA (Observe, Orient, Decide, Act) Loop

How do you make sure you are approaching the data in a clear, methodical way? We like the OODA loop as a best practice for making decisions within your Governance Body. The OODA loop is a concept developed by the U.S. military in the 1950s, initially used to describe how to react to the constantly changing air-to-air combat. Later it was expanded to encompass decision-making at various levels of business … tactical, operational, and strategic.

Steps of the OODA loop:

Observe: The decision cycle begins with capturing and aggregating data. For instance, collecting data from demographics, geography, competitors, and circumstances.
Orient: At this stage, data is analyzed and synthesized into insights, creating awareness.
Decide: After comprehending/understanding/analyzing the data, the course of action needs to be identified and the potential outcomes analyzed — solve problems, evaluate existing/new strategies, develop new hypotheses or testing plans.
Act: At the last stage, develop and assess the various options and then act on them. To simplify it further, implement solutions and test plans. The cycle repeats once the impact of the action is assessed.

Bottom Line

If you work to avoid these five pitfalls, your CDP implementation will begin to return impressive benefits for you right out of the gate. Ignore them at your peril!