– There is a debate about the effectiveness of using machine learning (ML) on first-party data and pushing it back server-side to Facebook to improve effectiveness.
– Many companies have raised money to solve attribution back to pre iOS14.5 levels and create audiences based on first-party browsing and clicking data.
– The customer journey is not linear and depends on various touchpoints, including audience, creative, homepage, landing page, offer, product page, organic channels, social presence, product itself, and price. It is important to pay attention to broader trends and collect clean data directly from customers to make better strategic decisions.Focus on collecting clean data directly from customers and leveraging that data to make better strategic decisions, rather than relying solely on ML and tracking scripts for attribution and audience targeting.We’ve been having an internal debate about the effectiveness of using machine learning (ML) on first-party data and pushing it back server-side to Facebook to increase the effectiveness of Facebook. There are quite a few companies that have raised money in the space in the last few months, all to solve attribution back to pre-iOS14.5 levels and help you create audiences based on first-party browsing and clicking data. Attribution has never been perfect though. It’s not so much the actions taken but the content that drives the actions, and they aren’t limited to just ads, which seems to be everyone’s focus. A single email, product page, landing page, offer, or creative doesn’t convert people. A customer journey depends on at least the following touchpoints: Audience, Creative, Homepage, Landing Page, Popup Offer, Product Page, Organic Channels, Social Presence, Product Itself, and Price.
The thing is, the customer journey isn’t linear, and it’s actually based a lot on outside sources coupled with one thing that is hard to understand: timing. It’s what Facebook did best because they aggregated your behavior across every part of the web to serve you the right ad at the right time. This worked because the data was plentiful and better than what any ad network has access to. This isn’t possible anymore. They took zero-party data too, such as your likes, shared posts, comments, etc., to create some really incredible models about behavior. So as I watch all these companies try to solve attribution, I start to question, does it even matter when most of them don’t have a solid experience on their landing pages, product pages, and are offering sales and chasing people with coupon questions? Not to mention, they are all using tracking scripts which, when they get flagged or big enough, are just going to be blocked at a browser level. Maybe I’m naïve, maybe this is the way forward, but I’d be willing to bet a lot more on the fact that rather than just installing a script and working off “cleaner” numbers, we all might be better off paying attention to broader trends, collecting data directly from customers, and leveraging that clean data towards making better strategic decisions. The truth is, most websites aren’t good. So that creative and its conversion rate, it’s not related to the creative, it’s a combination of the full journey. Change one part, affect the rest of it. Focusing on creative testing and more ads doesn’t matter if the rest of the journey has holes in it. But hey what’s a few thousand dollars a month for a quick fix that doesn’t address the underlying problems with your customer journey, it’s easier than actually giving a shit about your customers. At the end of the day, we’re putting a brace on a strained limb instead of building up strength around the injury to remove the need for the brace. We’re treating the symptom, not the underlying cause. #marketing #ecommerce #datahttps://www.linkedin.com/in/jivanco