It was a staple of the cartoons from my childhood: Seated on a river bank, an eager fishing enthusiast casts a line into the water and begins reeling in the line, imagining trout for dinner. Cue the laugh track — what breaks the surface of the water is a sodden old boot.

And so it is with marketing teams, enjoying the newfound freedom being pitched to them by various ad platforms. These platforms emphasize their ease of use in launching new campaigns. “You don’t have to wait for internally-generated Tracking Codes to deploy your marketing,” they say. “You can get the data you need with no hassle.” And marketers respond to it because it’s mostly true. The vast majority of campaign tracking codes are no longer generated by human analysts, but by the Facebooks and Doubleclicks of the world. Within their ecosystems, these platforms accurately track and monitor, dutifully feeding data into the tag manager.

But this presents a challenge to marketing analytics, one that can sneak up even when the tag management system is humming perfectly. When it comes time to analyze performance holistically, it works against your integrated marketing picture to have outside ad platforms creating cloned variations of codes that were carefully designed by the analytics team. Marketing teams don’t realize that in reaching for “freedom,” they’re also pulling in a lot of noise.

Ultimately, what marketing data is trying to tell you is, am I growing my business the right way? Am I spending money in the right or wrong places? The trend toward using automated campaign codes can only help provide that insight if it’s accompanied by another system, to grab the meaning of those tracking codes and map/upload those definitions into your analytics system. Your tag management system is capable of capturing endless quantities of data. The proliferation of automated cloned codes means you run the risk that meaningful insight is being pushed further away. Automation on one side (endless tracking codes firing into your tag management system) requires a complementary automation on the other side (telling your analytics system what those codes mean). Otherwise, someone still has to go in manually, pull and sort the data from these various platforms, and interpret their meaning for your analytics tool. The promise of Facebook and Doubleclick is that they will eliminate the manual handling of tracking code data. But if you want to integrate those data streams for a more complete picture, all they’re really doing is moving the manual data handling from the front end to the back end.

Marketers are human, and they often work against deadlines. They will take advantage of ad platforms that automate codes in the service of haste and expediency — not realizing that the meaning of that data may be lost. It isn’t their fault. They do it partly because they don’t (and shouldn’t need to) understand all the complexities of their reporting systems. As tracking codes increasingly get generated by outside platforms, teams will need to adjust by using better tools to relay those tracking code parameters to their analytics systems in such a way that the important pieces of metadata get captured.

What we all need is a system that gives us the best of both worlds: automation and clarity. A system that creates codes on a massive scale, and also automatically translates those codes into meaningful analytics reports. This would give our marketers real freedom to initiate campaigns, without dragging them through the minutiae of analytics implementations or risking that their reports will break. A system that runs itself, in other words, so that the analyst’s time isn’t spent on data triage. Until such a system is in place, the true insight of the marketing analytics endeavor (are we spending our ad dollars optimally) will remain elusive.