AI and Google Search: Putting Smartbots in the Kitchen
Yes, it’s AI hype time in the world of Search. And understandably, the focus is on the searcher experience, and how it will transform. It’s exciting, and a tad scary if you’re reading the news.
Less dramatic is a shadow realm behind the end user experience, a back kitchen where business owners and marketers toil, working to cook up effective ads in Google. That’s where I spend my day. And I can tell you, it needs some serious AI help.
So if we’re going to dream, let’s expand our vision beyond the AI-powered SERP, and dream of smartbots in the Google Ads kitchen. Close your eyes, take a deep breath, and imagine with me…
Smartbot sous-chef task: provide relevant, meaningful ad copy recommendations based on industry and performance.
As you may know, Google Ads likes to make recommendations. Among other things, it encourages you to use words in your ads that will attract clicks. (So kind, no?) It rates your ads - “poor”, “average”, “good”, and who doesn’t want to be good? It will even provide you sample headlines to consider. And many of these are completely worthless. Let me show you. Here’s a Google Ads recommendation to “make your headlines more unique”, with some very helpful inventory-related phrases. Small problem: the Google Ads account I’m in when I receive this advice is a service business. It has no inventory whatsoever.
Google Ads Recommended Headlines
C’mon, Google. We have no inventory.
For this same account, Google also recommends some shipping-related headlines. “We Ship Worldwide”, “Next Day Delivery”, etc. I’m sorry, but this is the definition of brainless. It’s simply providing a list, possibly based on some best practice, but without any additional intelligence.
Now, what if Google applied a little AI to this situation, and gave me headlines based on the business I was advertising. How hard can that be? Google already knows the business industry (via account information or by analyzing ads), so how about a smartbot that will give me tailored ad copy that maps to my, I dunno, NAICS code or something? And what if it uses its own performance data (gasp!) to customize the copy further, and recommend something to the tune of, “the top 25% of companies in your industry use headlines such as…”
I realize there are lots of implications here. Raising all advertising boats could result in lower click-through rates for the top performers, but isn’t better ad quality something to strive for? For the advertisers and the searchers? I think so. And taking this a natural step further, what if Google recommended specific offers and calls-to-action for advertisers? “The most successful advertisers in your industry offer…[fill in the blank -- 30 day trial, 25% discount, free design guide, etc.].” Wait, wait, you say, isn’t that your role as a search engine marketing consultant? Absolutely it is. But think of all the advertisers who would benefit if Google proactively offered this intel.
Smartbot sous-chef task: improve ad-serving diagnostics and bid strategy recommendations (and make Google more money in the process).
This one is a pet peeve of mine. A handful of times (one just recently), I’ve moved a Google Ads campaign to “smart bidding”, and it’s choked. Stopped serving ads. Zero impressions. I’ll observe for awhile. As I’m doing so, Google will provide helpful messages like, “Bid strategy learning”.
Bid strategy learning?
After 2 days of no impressions, how can it be learning? More like “bid algorithm broken”.
Sometimes I’ll watch for 48 hours, out of morbid curiosity. Or the distant hope that some software process will kick in and recognize that this campaign could be serving ads (and printing Google more cash). Hasn’t happened yet. So I’ll move bidding to enhanced CPC and boom, the campaign starts serving. This situation is ripe for AI. What if Google proactively analyzed my campaign and the account, and recommended a bid strategy that would serve ads from the get-go? Example: “We see you’d like this campaign to maximize conversions, but we don’t have sufficient conversion data to do this. For now, we recommend running maximize clicks (or whatever), and we’ll notify you when you can move to a smart bidding strategy that optimizes for conversions.”
And let’s get crazy, because we’re dreaming. What if Google could provide guidance along the lines of, “Advertisers in your industry tend to see a $50 cost per conversion, so you’re right on track.” Heck, PG&E lets me know how my energy usage compares to other similarly sized homes. Can Google out-AI PG&E? (Can a question be more rhetorical?)
Take several more giant’s strides down this path. Now I’m asking an AI master chef to build me a campaign to accomplish specific objectives, with ad copy that it creates, based on a snippets of unique value proposition that I provide. But let’s be honest. At this point I’m no longer in the kitchen. I’m in the dining room. Sipping a glass of syrah, marveling at the feast.
For those of you who witnessed my failed effort to use DALL-E to generate hearts with text on them, an update: DALL-E knocked it out of the park with instructions to create a “robot wearing an apron, standing in a kitchen”. I was blown away. And then things got a little creepy…
Need search engine marketing consulting? Or want to share your ideas on the intersection of AI and search? Drop me an email.
