Expand my Community achievements bar.

Tuesday tech Bytes: RTCDP - Week 06 – Maximizing Your Experience with Adobe RTCDP: Best Practices

Avatar

Community Advisor

8/12/24

Adobe Real-Time Customer Data Platform (RTCDP) is a powerful tool for marketers and data analysts alike. By enabling businesses to unify customer data from various sources and deliver personalized experiences in real-time, it has become an essential part of the modern marketing stack. However, to fully harness its potential, it’s important to follow best practices that ensure effective and efficient use of the platform. Below are some key strategies to get the most out of Adobe RT-CDP.

 

1. Establish Clear Objectives and KPIs

Before diving into the technical aspects of Adobe RTCDP, it’s crucial to define what you want to achieve. Whether your goal is to improve customer retention, increase cross-sell opportunities, or enhance customer engagement, having clear objectives will guide your use of the platform. Additionally, establish Key Performance Indicators (KPIs) that align with these objectives. These metrics will help you measure the success of your efforts and make data-driven decisions.

 

2. People to make it happen !

Many clients that we work with, end up spending big on the enterprise platforms but do not have the right people on the team. When you are out there shopping for a CDP, make sure you have the skillsets in the team to deliver on the vision the business has set out to achieve. This is why point number 1 above becomes even more relevant.

 

3. Prioritize Data Quality

The effectiveness of Adobe RTCDP relies heavily on the quality of the data you input. Ensure that your data sources are clean, accurate, and up to date. Implement data governance practices that include regular audits, validation checks, and standardization processes. Poor data quality can lead to inaccurate insights, ineffective personalization, and ultimately, a lower return on investment.

 

4. Be aware of the different guardrails for the system

We need to be aware of the different guard rails set for the system when we start out any CDP architecture design. You can find those guard rails here : https://experienceleague.adobe.com/en/docs/experience-platform/rtcdp/guardrails/overview

 

5. Try to keep an expiry to the Experience Event datasets

All the events that tie to a profile and adds to your license usage. You should keep an expiry for events that are not relevant to the business use cases. For instance, add to carts might not be relevant to be kept in the system beyond a certain time frame for the business. Similarly, there can be other data points which are not relevant post a certain time frame.

 

6. Do not enable any schemas for profile unless you’ve tested out the data model

There are instances where users enable schemas for profile without firming up the schema design properly. We need to keep in mind that once a schema is enabled for profile, there are no destructive changes allowed to the schema. What ends up happening in this instance is that if you’ve enabled a schema for profile prematurely, you end up creating multiple ‘versions’ of the same fields or different ‘versions’ of the same schema.

 

7. Take the velocity of data into account when thinking about audiences

A lot of users think of streaming audiences when looking at their use cases BUT the data attributes or events that is based on is a batch ingestion. The velocity of data should align with the audience definitions.

 

8. Have a thorough QA of any data ingestion - Especially identity related information

Since RTCDP is great with identity stitching, it becomes imperative to QA any data ingestion and ensure that no identity fields have static values or dummy values being ingested. If this happens, you would end up stitching multiple profiles into one, popularly know as, a profile collapse.

 

9. Put into place a robust data governance strategy in place

Since a CDP does contain PII, who gets access and to what information within the platform is important. It is also very important to label your data fields as far as possible with the relevant labels. Having a policy based on the marketing actions based on the labels, makes sure that you, as a business, are abiding to the contract with the end user. @Danny-Miller has put together a must-read article around Data Governance which you can read here: https://experienceleaguecommunities.adobe.com/t5/adobe-experience-platform-blogs/getting-started-wit...

 

10. Make use of upsert feature to avoid overwriting profile information

As an architect, make sure you've thought through the data ingestion so that you do not end up overwriting information. At times, when you have a dataset which ingests data from multiple data sources during the day, you might end up overwriting the profile information. Make sure to take into account such a scenario and enable upsert for the dataset if you think that this might be relevant in your case. Here's a link where you can read more about the upsert functionality: https://experienceleague.adobe.com/en/docs/experience-platform/catalog/datasets/enable-upsert

 

Hope the above pointers got you thinking !

 

That's a wrap to the Tech Bytes series ! If you would like to read the earlier blogs, below are the links to the earlier posts in the series.

 

Week 5: https://experienceleaguecommunities.adobe.com/t5/real-time-customer-data-platform/tuesday-tech-bytes...

Week 4: https://experienceleaguecommunities.adobe.com/t5/real-time-customer-data-platform/tuesday-tech-bytes...

Week 3: https://experienceleaguecommunities.adobe.com/t5/real-time-customer-data-platform/tuesday-tech-bytes...

Week 2: https://experienceleaguecommunities.adobe.com/t5/real-time-customer-data-platform/tuesday-tech-bytes...

Week 1: https://experienceleaguecommunities.adobe.com/t5/real-time-customer-data-platform/tuesday-tech-bytes...