Master customer data with a modern approach

When considering the future direction of their martech stack, many organizations are quick to adopt a modern data stack as the foundation. They are anchored in cloud data warehouses and use tools and technologies that are extensible and interoperable within a composable architecture.

The benefits of a modern data stack are legitimate and far-reaching, including speed to market, agility, and reduced costs through limited operations. However, even a modern data stack for customer data will not deliver the expected value if you do not adopt a modern approach to using customer data.

“Customer 360” is misunderstood

Everyone says so and everyone seems to want it. There are too many martech platforms and organizational initiatives with “360” in their names. For too long, we have been led to believe that personalization is the primary way to drive growth, and that a 360-degree view of the customer is the best way to enable that.

Customer 360 is a great goal, but an irresponsible pursuit. Not only would it take too much time and money to accomplish, not to mention the opportunity cost of missed growth opportunities that could have arisen by working harder.

With the right customer data strategy, chasing data can be a distraction and keep you away from Customer 360. There's too much to learn. The best customer data strategy is based on learning challenges. Identifying clear learning objectives that align with achieving measurable business outcomes will guide you along the way. If planned properly, this approach should prevent him from realizing his dream of a 360-degree customer view. And the results you drive will make you wonder why you had that dream in the first place.

We recently found that this approach was able to convert all the leads generated through an upper-funnel strategy at a travel company with limited sales department capacity. They focused on understanding booking trends rather than working on understanding a more complete profile of previous guests and past interactions. They redesigned their lead scoring model using machine learning (ML) based on fundamentals such as customer profiles, web interactions, and conversion data.

This approach has replaced the traditional overly complex approach to lead scoring within marketing automation platforms. This allows them to invest their sales efforts more effectively, improving operational efficiency and increasing bookings.

Learn more: Real-time customer data platforms: promise and reality

Build Customer 101 first

The College 101 Level Course is an introductory course that focuses on the basics. The approach to building a customer data store should be similar, solving basic data needs.

use right data

Forget about having all the data about your customers from every interaction with your brand across all channels. Appropriate data includes:

• Key customer profile characteristics.
• Interactions from the most common touchpoints.
• The actions most associated with the most important metrics that drive growth (i.e. new leads, repeat purchases) or hinder growth (i.e. churn).

Best practices for data management

Establish core functionality as part of your foundation. Implement good practices for data cataloging, data governance, data lineage, and data privacy from day one.

Create better data

Using the right data is only as effective as the data is good. Ensuring proactive and ongoing data quality practices are applied to your data environment is paramount to ensuring data consistency, availability, and most importantly, reliability.

Let's dig deeper: Customer data debt: The hidden barrier to CDP success

Resolving identity is key

A key assumption of a customer data store is that it resolves a single customer view (SCV). SCVs cannot be created without applying identity resolution to connect and dedupe customer data sources. The level of sophistication required depends largely on the nature of the data.

• How and where customers identify themselves to the organization.
• Different identifiers across data sources (terrestrial vs. digital, email address only, etc.).
• Major PII gaps exist within your data sources.
• Level of data standardization and cleanliness.

Identity resolution (IDR) can be complex, and market providers are similarly diverse.

IDR 101

No matter what approach you take to resolve IDs, your solution must handle the basics and be able to:

• Construct and persist an identity graph.
• Adjust and expand your identity graph and matching rules as you add new sources and identifiers.
• Generate and assign persistent IDs to uniquely identify individuals/users/customers.

There are multiple approaches to resolving IDR. Examples include leveraging ID Spine providers, MDM platforms, packaged CDPs, or even deploying your own ID graph as part of a custom solution.

101 to enrich your customers

Rather than consolidating all your customer data for a Customer 360 view, find creative uses to enrich your Customer 101 dataset through an enrichment approach. The best customer data strategy is based on a data-rich learning agenda.

Start by identifying ways to enrich your core customer data assets, including:

• Develop a predictive model using machine learning to generate new segments. Examples include purchase trends, churn trends, next purchases, and channel trends.
• Analyze data and generate new customer segments (i.e., high-value customers).
• Build lookalike models based on high-value customers to expand your prospecting reach.
• Leverage second-party data by sharing it with your partners.
• Overlay customer data with third-party data sources such as demographics, psychographics, company statistics, weather, and more.
• Use publicly available datasets such as climate and environment, census, economic, geospatial, transportation, and travel.

use, prove, move

After introducing new capabilities such as a CDP or unified customer data store, organizations often seek to restructure many of their previously adopted customer experience strategies. They often lack evidence that past strategies are the best approach.

New approaches cannot be tried due to the potential risk of negatively impacting traditional KPIs. However, if you take an immediate and deliberate approach to testing new strategies, you can quickly discover new insights from your learning tasks.

Employing a continuous improvement process allows you to quickly iterate on each learning objective. As shown below, you first create an enrichment, then adjust your program to leverage that enrichment in your tests, and derive insights from your test results and performance metrics.

Use enrichment strategies applied to your data

A common first step is to better understand who your customers are and therefore which customers are of high value. Which ones are most interested in your brand? To learn this, start by enriching your good data with things like machine learning models.

Generate new segments by leveraging the model's scores. These segments should be deployed and made available to the user community in areas where they consume consumer data.

• As a dimension in customer reports and campaign performance dashboards.
• As a segment used anywhere you build an audience, such as in a CDP.
• Within the product, as a user attribute or a dynamic value.

prove (or disprove) a hypothesis

Get your measurement and testing framework in place. This means creating a structured iterative learning methodology that allows for quantifiable measurements. Please check the following:

• Define and align your business metrics.
• Document the formula for each metric to increase visibility for your team.
• Organize your metrics into separate categories.
• Utilize experimental design to achieve actionable measurement results.
• Create a statistically significant sample size for your test.

Whether you apply the insights directly or test them first, it's important to plan your measurements to ensure you can quantify the impact.

Go to next learning objective

As you derive insights from your tests, make any necessary adjustments to your strategy based on those learnings.

• Adjust your channel strategy to optimize the experience you're creating.
• Tune your ML model with additional features.
• Reduce marketing waste to unresponsive audiences.

In some cases, learning is complete and you can move on to the next learning objective. Perhaps you have learned that you need to investigate your initial hypothesis further. Or, it's time to move on to the next learning objective on your learning agenda. Follow the breadcrumbs of data.

A typical development begins with the question, “Who is our customer?” That question then leads to a series of questions such as:

• Who are our high value (HV) customers?
  • Who is most likely to have HV?
    • How can you get more HV customers?
      • How can you get more HV customers?
    • Who is likely to buy next?
      • Which products are HV customers likely to purchase?
        • Which product is each customer most likely to purchase?
• Which customers are most interested?
  • How does their LTV compare to other cohorts?
    • Which products are you most likely to buy?

Each of the above questions requires another cycle of continuous improvement to build enrichments, test in the market, evaluate results and move on to the next idea, as shown below.

Dig deeper: How to categorize customer data for actionable insights

Rethink your customer data strategy

There isn't a team that doesn't feel the pressure to get more from less. Modern approaches to evolving data assets require getting back to basics and avoiding distractions from the nirvana of a 360-degree view. There's no harm in working on creating 360-degree views, but if this is the core of your data strategy, you may be missing out on other, more valuable, and more creative opportunities along the way.

Taking a creative approach to the use of data includes enrichment, but also other ways to monetize customer data, such as enabling other corporate functions such as sales, service, and even finance. I will also look for ways. A modern approach to customer data means establishing customer data as a core function within your organization, rather than an ongoing project.