SGE Ads Are Inevitable. What Do We Need to Know?

Search generative experience (SGE) is coming to Google – it’s currently in beta and evolving on a daily basis. When it rolls out fully is anyone’s guess, but it’ll likely be a matter of months, not years. And where the Google SERPs go, ads will follow.

Search Engine Land posted about this recently, showing a test of sponsored content ads that popped up in SGE as a response to follow-up questions. Google will certainly run more tests of ad types and different UX placements before rolling anything out in full for SGE, which itself hasn’t launched yet. In the meantime, it’s worth considering how those ad types might work, and for what kinds of brands.

6 Early signs point to eCommerce usage

Ultimately, if SGE ads are effective (read: they make Google money), they’ll be more or less universal. However, I think eCommerce is going to be affected sooner than B2B for a few reasons:

1. The Google labs SGE landing page showcases personal lifestyle prompts and shopping among the main ways to use SGE (i.e. “I want to see a range of products and things to consider while shopping.”)
2. Google recently released a Generative AI Gift Guide that points toward eCommerce sites and products.
3. The photo in Search Engine Land’s article showcased a listicle about the ‘Top 8 Fly Traps.’ 
4. Considering how many product review listicles are published (especially during gift-giving seasons), there’s a ton of potential eCommerce content the algo can pull from.
5. Google has been slower to innovate ad types for B2B in recent years; even Performance Max campaigns aren’t mandatory for B2B yet, presumably because they’re geared toward eCommerce/B2C.
6. B2B’s relatively long journey to purchase is kind of a testing double whammy. Advertisers would be hesitant to explore a new ad type that might or might not work, and Google wouldn’t get quick returns on its effectiveness.

Machine learning and impacts on performance might be cause for concern

Ads served in SGE will be based on two kinds of machine learning: 1) the algo driving SGE results; 2) the algo Google uses for ad targeting. With two potentially sub-optimal algos in play, the odds of poor irrelevant targeting go up pretty significantly. 

More of an immediate concern for advertisers: it’s already expected that SGE will lower organic and paid CTR. This stands to lead to a hit to market share and revenue.

The additional layer of qualification (the initial prompt a user submits) also raises the question: how broad do we have to get? For example, how would advertisers be able to use exact-match keywords in SGE? Broad and phrase match seem to be far more applicable in SGE scenarios – to the extent that keyword-based bidding will be an option at all – which may or may not be in line with a business’s strategic goals.

One possible UX scenario: 

1) The SGE algo scans a website’s content and qualifies the advertiser. If a user asks, “What sound speaker is best for a pool party?” the product on a business’ site needs to match.

2) The targeting algo comes into play ONLY IF an advertiser’s website qualifies.

It might be an exercise in futility to guess how this will play out, but I can see where selecting prospecting audience segments might be a good fit, specifically for targeting in-market users or users showing affinity, or just casting a net for broader audiences to see what works.

Taking it a little further, retargeting might be a bit trickier. Advertisers would potentially need to use a Website Visitors audience to retarget the SGE users in another ad type (search, display, Performance Max, etc.), and since everybody’s SGE experience is tailored, it would be challenging to retarget them based on the initial prompt.

Ultimately, of course, our recommendation is to keep a close eye on news and developments as Google rolls out more tests. If you’re curious about possible betas, make sure to let your Google rep (or your agency) know that you’re looking to get early access. And keep an eye on our blog for updates!