– iOS 14 prevents Facebook from tracking users across websites for targeted ads
– Unique coupon codes are a more accurate way of measuring attribution than email addresses
– Understanding correlation and causation is crucial in marketing and data interpretationCollect, interpret, and leverage data effectively to understand correlation and causation in marketing.Friendly reminder: iOS 14 prevents Facebook from tracking you across every website you use and using that information to get a better idea of your interests in order to serve you ads. Although it impacted attribution reporting, that was the most minor impact of the change. On that note, the most accurate attribution you have is via unique coupon codes. You could do this super simply by just allowing people to unlock a coupon code by pushing a button rather than requiring an email address for it. I wouldn’t recommend it, but you could do it. The best performing ads typically have the highest opt-in to conversion rate for subscribers to unlock a discount, and further on that point, it’s not about ROAS but about the number of orders. ROAS, by definition, is “Return on Ad Spend,” which means revenue collected versus what was spent on ads. It does not factor in repeat purchase rates, differences in product prices, profit margins, etc. For that reason, it also cannot be measured 1 to 1 in a time period. Positive ROAS on low orders usually means a lower repurchase rate depending on the products purchased (higher value). This is correlation and causation in a nutshell. A lot of companies are following the wrong intent signals. A lot of the tech those companies are using is highlighting the wrong intent signals too. Not all pieces of data are created equal; data without context is the leading problem we have in marketing. Knowing a subscription to conversion rate becomes a lot easier when you understand the relevance of why someone subscribed. Yet very few collect data during a signup to qualify the subscription. Correlation and causation are the two biggest misunderstood parts of marketing; we see things that don’t exist and create relationships that are often caused by unrelated events. Most marketers would also do well to take a statistics and statistical modeling class. It seems like the most prominent tech companies are all pitching narratives that can’t actually be backed up via data. Remember, it’s all about data interpretation. Most ecommerce companies would be better off understanding the unit economics of running ads, fine-tuning the journey, putting together an irresistible offer, and getting the product in hands at all costs. Profitably. A store only becomes a brand when enough people know about it and enough product is in the wild. Marketing technology, including our own, requires you to understand correlation and causation and be smart enough to recognize the difference. There are no cheat codes for this. It’s only a matter of understanding how to collect, interpret, and leverage data. Here’s the difference though, at Formtoro, we’re playing with a larger data set, a cleaner data set, an intent based data set that is closest to a transaction. I have less things to question in our data than other platforms. This is on purpose.https://www.linkedin.com/in/jivanco