Step 4 - Audience Segmentation
The fourth step is to organize the profile data and segment into different audiences for analytics and marketing. Common audiences include first-time buyers, high-value customers, customers who are at risk of churning, or members of the loyalty program. The right historical and predictive attributes are a must to drive segmentation — examples include up-to-date historical and predictive lifetime value, preferred product or preferred channel, and an up-to-date view of contactability, including opt-in status.
It is important that this segmentation capability allow for rapid exploration that can be used for measurement (growth in loyalty signups), analytics (impact of campaign activity), and marketing (opportunity sizing). Segmentation needs to scale to serve sophisticated SQL users as well as marketing and business users, powering both analytics and marketing. Any segmentation approach must fit in with the data infrastructure of the brand — including rich connections to marketing clouds and data warehouse platforms.
Step 5 - Orchestration & Distribution
The fifth step is to distribute (or orchestrate) the customer data to all of the downstream systems used across the brand for analysis, engagement, business operations, and insight. Companies have invested years in training marketers and analysts on world-class tools — from data visualization to campaign management to customer experience — and there is no need to replace these systems. Instead, a customer data engine powers them with better data. The engine needs to power the full distribution of the customer master, as well as lists and up-to-date attributes, across all systems in the brand.
Step 6 - Measure, Learn, & Iterate
Last, but possibly most important, the combined customer data systems must support rapid measurement, learning, and iteration. This starts with the ingestion of all engagements and interactions in real time to power up-to-date metrics about customer and business health, including the ability to measure the results of campaigns or actions being taken.
This means the engine needs to support a rapid iteration from Step 1 through Step 5. An equally important part is the ability to adapt the systems to changes in the business — new data sources coming online, acquisitions and mergers bringing companies together, and new tools coming to market. A brand’s network of customer data systems can no longer just be resilient to change; it must embrace it.
Putting It All Together
These six broad steps describe capabilities that are required in any brand ecosystem to drive personalization at scale. And they must fit in with the technology stack of the brand, so technology leaders require solutions that meet them where they are. In some cases, brands will want a customer data platform (CDP) that powers all six steps. In other cases, brands will choose to use a marketing cloud for segmentation and audience creation, while picking a CDP for data collection, identity, and profile creation. And in other cases brands will collect data in a data warehouse and then use a CDP for identity extraction, profile creation, and orchestration throughout their ecosystem.
This is one of the reasons we built Amperity with the flexibility to power any or all of these stages that drive transformation, and is what allows us to stay laser-focused working with technology leaders at brands on accelerating transformation that builds from where they are.
To find out more about how Amperity can help you follow these six steps to personalization, get in touch.