We've all been there with campaign management — marketing and analytics teams have a 40 billion-person audience in mind, and the delivery team has the campaign primed and ready at 2,000.
Traditionally, we’d spend endless hours combing through each segmentation filter only to discover that data is not fully present for each data point. So campaign filtering would get dropped, and that finely tuned 121-omnichannel campaign would look more and more like a “one-size-fits-all” batch-and-blast to the moon. Accusing fingers would point to upstream data ingestion, and tickets would sit across the business, only to be forgotten until the next big launch campaign.
Today, Customer Data Platforms (CDPs) let us access all our data in one place. Fantastic!
Or is it?
Because now there’s a sea of red error reports for incoming data, and the accusing fingers are pointing at the CDP itself.
Worst of all, no one is owning this. And by the way, who even owns this? Who could own it? Who wants to own it?
Anyone?
Bueller?
But all’s not lost. Like so many things, you’ll see better results with customization. In this article, I’ll share how applying best practices around your data will yield big benefits for your CDP.
The Good Old Days: No Data Monitoring Meant Ignorance Was Bliss
Before CDPs were widely used, customer data flowed around a spider web of different platforms, and each connection had its own integration.
The majority of web personalization platforms have no data monitoring. If there’s a front-end script running that uses website meta tags but the script fails based on a page not having a tag or a particular syntax within the tag, data corruption occurs. Back in the day, no one would ever notice such edge cases, because the bad experience only happened on the customer’s browser.
CDPs now act as the centralized data system that sees all data flowing between end users and backend systems. With greater awareness around data, there is now more to manage.
Today's Solution: Consolidate CDP Reporting into Something Actionable
Most MarTech platforms are just foundations. The true magic of the user experience is made possible by building custom solutions on top of said platforms.
That’s true of CDPs as well. They are not the root cause of poor data, and they do have the ability to provide a solution — they just need some light modification, because the current reporting is not effective.
Let’s use an example from Adobe Experience Platform’s Real-Time CDP. When looking at the datasets view above, it might appear that all our incoming data is golden, because this dashboard only reports on the most recent data ingestion.
However, when we look deeper into a given dataset, we can see a ton of failed data loads:
Most clients average around 200 datasets, so going into each individual report is not feasible. We need to consolidate this reporting into manageable information.
Fortunately, Adobe CDP, like many CDPs, provides an API service that can supply all the information we need on dataset ingestions, including status, error codes, creation date, and number of records. Here's a view from Adobe RTCDP's Catalog Service API:
Using something like Adobe Runtime, Google Cloud Services, or AWS Lambda, a script can be written to:
- Call the RTCDP Catalog Service API
- Read the API response
- Carry out any necessary data transformation
- Aggregate our data accordingly
- Export data to a dashboard
- Send notifications via Slack or email
With a single snapshot of all data ingestions, we can prioritize resources accordingly. (In terms of dashboards, I’d recommend using either an existing client service such as Google Big Query or Power BI, or ... just building a custom dashboard.) Let's take a look at the data in our example:
Here, we can see a recent drop in website data quality, and remedying it will be an urgent task. With even more analysis, we can dig deeper, refine, and improve the customer experience.
Embracing Data Diligence
As marketing organizations integrate CDPs into their MarTech stacks, the technology will evolve in parallel with the teams using them. It does, after all, come with its own set of challenges. However, I believe we are headed in the right direction — it just takes a little more discipline around data and some basic customization to find the full potential of your CDP right now.
If you're looking to implement a CDP or need help managing an existing platform, reach out to our experts.
