The Algorithm Isn't As Powerful As You Think

Ad platforms like Google and Meta have come a long way with the development of their machine learning algorithms. They’ve provided advertisers with great features like automated bidding and optimized targeting.

While features like broad / advantage+ targeting on Meta are certainly powerful, it can be difficult to drive sustainable success using these tactics in a silo. To be more specific, even DTC advertisers may struggle to drive efficient performance running only advantage+ shopping campaigns or performance max without supplemental marketing efforts.

One example here is the recent shift in strategy from Nike. Over the course of the past few quarters the company doubled down on “performance” marketing by leaning into bottom of funnel efforts like those I mentioned above. This came at the cost of some of their traditional branded efforts.

The result; as of the end of July Nike’s stock was down 30% YTD.

My big takeaway here is that even with millions (or even billions) of data points to feed these algorithms, Nike’s campaigns started to deliver lackluster results. It appears that even in the DTC world, people will still conduct their research and go through a consideration phase before making a purchase.

To this point, as great as some of these algorithms are at finding people who are most likely to purchase, no algorithm in the world can actually force someone to take an action. The key word in that statement is most likely. The algorithms developed by Google and Meta can find people who are the closest thing to other customers, but sometimes even the closest Instagram user hasn’t even considered buying a pair of Nike shoes recently.

Relying on an algorithm to do all of the heavy lifting may not work out for advertisers. It didn’t for Nike.

That’s why I recommend all of my clients explore some sort of complimentary brand/awareness/demand gen play to keep their brand top of mind for when it does come time for a potential customer to make a purchase.

Have questions, considerations, or critiques? I’d love to hear them! If you’re reading this via email, just hit respond. Otherwise, you can find me on LinkedIn.