As Google Chrome is sent to send Cookies to the dustbin this year, digital marketers will lose one tracking mechanism on which they have come to rely…perhaps too much so.
A primary application for browser cookies in marketing is the ability to remarket/retarget a browser (person) that has been to the website. This seems great, as it shows interest. But, it has led to lazy marketing habits.
Blanket remarketing gives no real thought to context or timing. It intrudes on users when they are completely removed from the context or concerns that led them to the advertiser in the first place.
But doesn’t Google automation deal with this? Not really. In theory, it should. But we’ve all experienced the unsettling sensation of being “followed” around the Internet by a site we visited.
This recent article in AdAge speaks to this very issue.
The next level of automation
As Google pushes everyone to trust their AI, it expands on the concern raised by blindly retargeting users based on cookies: that, as marketers, we are relinquishing our control to a black box of automation. We are told this improves our marketing ROAS, but I am not convinced.
I think AI will improve results when it comes to unmanaged programs. For marketers who do not have the time to monitor and adjust their campaigns, AI provides a level of optimization that is bound to make things better.
I have seen dedicated campaign managers do much better for the average company. One thing AI is good at is finding the low-hanging fruit. A freshly launched program managed by AI can get quick wins, but they often plateau. Changes to pages, offerings, and budgets can reset the AI to a learning phase. Manually managed programs can get beyond the low-hanging fruit.
The other area where AI will excel is data analysis and response. However, the response/output is only as good as the data input, and the average company has issues with this.
A full circle of ads, outcomes, and feedback works well for AI. The most basic of these is E-commerce sites that feed all the sales back into Google Ads (directly or through G4.) AI is likely your best bet if you have an e-commerce website, and your primary conversion is an online sale.
Where Google AI falls short is when the value of an action is determined after the conversion point. All Google knows is that a conversion point was triggered. Herein lies the problem.
Left to Google, the initial round of leads is more likely to be filled with junk than value. However, Google doesn’t know this, so junk leads are used equally in conversions. The AI is optimizing the program to deliver junk.
Can this be fixed? Yes. Absolutely. But not by the resources of the average company.
Fixing Junk Leads in Google Ads
Google provides valuable mechanisms to improve lead generation for offline activity following an online conversion.
Every lead submitted to Google Ads should include the corresponding lead ID, which should be carried with the lead through the sales/closing process.
When leads are closed, a file can be generated (or API used) to load the results back to Google. With this file, Google can assign value to each lead, allowing the AI to optimize lead generation.
For junk leads, which are usually identifiable very quickly, a file with the lead ID can be uploaded to Google to have these leads deleted from the system.
Like so much of the ad tech hype, leading-edge tools require resources and knowledge that are not available in the average company.
Beyond the issue of resources is that of reliance on technology and complacency. Remarketing as we know it will end this year. And that’s not a bad thing. It reminds us that we are prone to accepting the easy out, allowing technology to do what we should be doing—or at least paying a great deal more attention to.
As we continue to push the envelope of AI in advertising, we must always pay close attention to the outputs and ask ourselves if this really is the best that can be done.