AI and SEM - True Love?

On this Valentine’s Day, I thought we might contemplate the union of artificial intelligence and search, particularly search engine marketing and ads. I must admit, and I’m not proud of this, that my first reaction to their conjunction was anxiety. I read a NY Times article about the threat that AI posed to Google search, and I panicked. I sold my Google stock. I started planning for a new career. (I’m joking. Half joking.)

Later, after contemplation and a couple podcasts, I am optimistic. This could be a legitimate romance.

Looking at Bing’s sample AI + search results pages helped me turn the corner. Bing appears to be retaining the traditional list of websites, while introducing an “AI chat window” on the right hand side (blue box below). I looked through all their sample pages, and in many I see ads being served in the SERP list. (Whew, not out of job quite yet.)

What will be interesting/fascinating to witness will be the interaction between the AI and search ad serving. Will I be able to tell Bing to show me ads for specific products? (On second thought, will anyone do that who’s not a marketer?) More plausibly: if I’m ready to buy, will the AI be able give me a list of products that fit my criteria, and will that list be created, at least in part, with ad data? How cool would that be?

Quick illustration: I bought speakers recently. For my living room, to sit on either side of the TV. The process took weeks. Months. I read product review articles. I watched YouTube videos. I googled like crazy. I finally decided, and bought them. They arrived - no, wait, Amazon said they’d arrived, but they hadn’t! I waited. I stewed. I requested a refund. I bought them again, this time from the manufacturer because they required a signature upon delivery.

Now imagine how much more efficient that process could be, with a lot of AI help. Here I am chatting with Bing…

“Give a list of top-rated speakers”

“Show me the ones that are less than 34” tall”

“Passive only, black finish, 5” woofer”

“Less than $500 each”

“Are any on sale?”

“Does delivery require a signature?”

If Bing could respond by tailoring the SERP with each instruction, or better yet, simply serving me the URLs in the chat window, that would be a massive timesaver. Something like this requires a ton of computing power, which is one of AI’s obvious challenges, but it also requires very detailed product information. Someone on the product side will need to supply this data for the AI to use. Will this data come from the product manufacturer’s website, or from what’s provided to Microsoft Advertising, or some other source, or all of these? Who knows. We’ll see. (I’m still waiting for a golden ticket.) But the potential to speed the decision-making and the movement to buying is tremendous.


As a funny AI side note, I initially tried DALL-E to create the lead image for this post. How appropriate to use an AI image generator, yes? And how easy it will be for DALL-E, to simply impose a string of characters - AI+SEM=TL - over an image, right? Well, after 30 minutes of text inputs and 40 heart images later, I gave up. (Actually, then I tried Craiyon. Nope.) Sigh. The dream endures…

Previous
Previous

AI and Google Search: It’s Hard to Kick the Click Habit

Next
Next

Yes, Bing, We Are Waiting…