Automated Data Systems: Why the Human Touch and Quality Control Matter

Posted on by Beth Negus Viveiros

Faced with tighter schedules, more information than ever to manage and greater expectations from upper management, many marketers are turning to automated data-marketing systems.

There are plenty of transactional/trigger-based systems and features that automate and optimize data flows based on transactional, lifecycle, event-based, trigger, behavioral, and demographic variables.

But even the most sophisticated systems are only as good as the people, teams and processes overseeing the process. Here's five key points you should keep in mind.

1. Make sure everyone is clear on the business goals of the project.
Everyone on the team should have a clear understanding of the system and the business goals. This includes the technical team members responsible for programming the final logic flow such as getting live feeds into the system, sending out confirmation or welcome messages, and so on. A common mistake marketers make is sending tasks to their fellow technical team members piecemeal instead of explaining the big picture. The more complicated and sophisticated the system, the more important it is that every team member has the overall view to guide them with their individual parts.

2. Have a quality assurance (QA) plan and map at the outset.
Programming and setting up a transactional message system correctly is key. This is especially true for large online marketers with many data input points, hundreds of logic operations, and exponential amount of data flows; there needs to a common map to implement the data flows and test it on an ongoing basis. Marketers are so often in hurry to get it "up and running" that they lose track of the goals, and the name system becomes a misnomer for a patchwork of sub-optimized data flows.

3. Designate a project owner.
There needs to be someone to bridge both the marketing and technical implementation details of the project. For instance, if a live feed needs to be set up from a sign-up site and push out welcome message X, then someone has to make sure that this has been done correctly. That is, that the right live feed and parameters were set up with the right matching creative.

4. Implement a change process.
The system may be set up and working according to the requirements and business goals, but it doesn't exist in a vacuum. Software upgrades, database tweaks, domain changes, new features requests and so on all run the risk of breaking something.

Ironically, it is not the serious problems that one has to worry about the most because these will readily show themselves. Rather, it is the subtle, easily missed break that can cause long-term problems until it is detected from a review. Examples here are the image that is no longer being pasted in the creative reducing the subscriber's experience (and engagement), or a delay in sending out the subscription confirmation email that reduces conversions.

The point is that a change-system process has to be put in place to ensure the integrity of the system and to signal the need to re-test the system data flows according to the defined QA map.

5. Test, test, optimize.
Testing before going live—and periodically while live—is important. Too often everything will perform well in the pre-test environment when the load and demands on the system are low. But it becomes a whole different environment when the system goes live; after all, the load on the database server, web server, and message transfer agent all increase where one weak link could cause problems throughout the system.

Good planning and a QA plan at the start will minimize these problems, but it is nevertheless important to systematically and routinely test the system on an ongoing basis.

In addition to detecting problems, a systematic testing and monitoring system should also be put in place to optimize data flows. The most one can hope for at the start is a good system, but with the right testing and monitoring this can be improved to a great system.

While successful marketing increasingly requires more sophisticated analysis, tools, and systems, the management of teams, projects, and processes becomes even more important. Data-marketing flows can be automated, but good planning, management, and coordination cannot.

Andrew O'Halloran is chief privacy officer and manager of industry relations at Cypra Media

Related articles:


Related Posts

Chief Marketer Videos

by Chief Marketer Staff

“Trapped in Advertising” sounds like a hashtag you might use after a bad day at work, but for Pringles and Adult Swim, it was a year-long, pop culture-forward campaign aimed at ad-averse 18- to 35-year-old “Rick and Morty” fans. The goal was to sustain engagement and become part of the cultural conversation in order to…


PRO Awards


Click here to view the 2021 winners.
Information on 2022 coming soon!

CHIEF MARKETER 200 Winners Announced

CM 200

Click here to to view the full list of the 2022 winners!