How to Future-Proof Your SEO Strategy: Decoding the True State of Generative AI SEO with Expert Eli Schwartz.

Editor’s Note: This article first appeared in Kyle Poyar’s newsletter, Growth Unhinged, which explores the unexpected behind today’s fastest-growing startups. You can subscribe to get the latest from Kyle here.

Eli Schwartz literally wrote the book on scaling organic acquisition of PLG products – Product-Led SEO – and he helps companies use SEO to unlock billions in revenue. But before all that, from 2012 to 2019, he headed up global SEO at SurveyMonkey, where he launched the company’s very first SEO programs – an incredibly lucrative initiative that drove two-thirds of the company’s global revenue.

I chatted with Eli recently about what generative AI (gen AI) means for SEO. This isn’t another one of those “SEO is dead” think pieces; it’s an expert’s view on how to navigate changing buyer behavior. Eli shares his insights on the current health of SEO, explains the massive impact he expects Google’s gen–AI powered search – Google Search Generative Experience (aka “SGE”) – to have on SEO and offers up his advice on what you can do to win at SEO in 2024.

Here’s Eli, in his own words.

SEO isn’t dying – but it is morphing into something brand-new.

I don’t think SEO could ever possibly, truly die as long as people need to find and fetch information online from external sources.

But let’s face it. SEO as it exists today isn’t a great channel for users. There’s too much noise, too many companies created solely for SEO purposes. Search for “best credit cards,” for example – you’ll see most of the results are content sites that make money from affiliates. It’s not a great experience, but it’s how SEO works right now.

So I believe SEO is going to be reborn. Gen AI will absolutely change what SEO means. Gen AI will make a lot of last year’s SEO best practices obsolete. And gen AI will shift the focus of future SEO efforts to an entirely different place in the user’s journey.

Take Google’s featured snippets – those extended search results that appear at the top of Google. They give users a quick answer to their question right in the search results by extracting information from a web page that’s already written.

Google extracts information from a web page that’s already written to create a “featured snippet.”

But Google SGE – or ChatGPT or whatever comes next – on the other hand, will write the content it presents to users on-the-fly, after it extracts and analyzes all the information it can find related to the question asked.

You can already see the experimental version of SGE on Google right now, and it’s pretty compelling.

Google’s gen AI search, on the other hand, extracts information from all the search results it finds, then writes and presents an answer to the query.

With the new way that gen AI SEO works, links and backlinks and domain authority won’t matter much anymore. But content – the richness of the content, how that content is laid out, etc. – will.

And that means how you use keywords within content and how you structure content will change. Whether you use synonyms, plurals, the foreign-language versions of words – all of that will change, too.

We’re going to move into this new, gen AI search world where, for example, you’ll just be able to ask “what are the best credit cards?” and Google will tell you, based on its large language model (LLM), which five credit cards are the best ones for   – and it’s not unlikely that Google will even personalize its suggestions. You won’t have to actually go to the credit card websites to extrapolate the answer.

“You’re gonna need a bigger boat.”

All of this is why, if you want to do SEO effectively in 2024, you can’t focus on top-of-funnel (TOFU) alone anymore. You’ll have to focus your SEO on a completely different place in the user’s journey.

That’s because SGE/ChatGPT will answer the quick, TOFU search questions, so your SEO doesn’t have to – questions like the ones below will already be handled for you:

• Which is the best solution for [abc]?
  • How to implement [abc]?
  • What is [abc]?
  • Symptoms of [abc]?
  • Reviews of [abc]?

A generic, top-of-funnel, gen AI search response.

And another…

Instead, you’ll have to build keywords around the value you deliver mid-funnel (MOFU).

Let me say that again: You’ll have to shift your SEO efforts away from TOFU users and redirect them toward MOFU users.

And that means you’ll have to do a much, MUCH better job of understanding your customers.

Right now, most SEO pros don’t do enough research on the customer beforehand and just consider keyword research to be all they need to know about the user. But effective SEO means deeply understanding the buyer and the buyer’s journey, then knowing enough about search to know where your results should fit.

Going back to my credit card example, the same number of credit card applications will be submitted this year. But in the SGE/ChatGPT world, it won’t be enough to show up at the top of the SERPs for TOFU searches. You’ll need to move your efforts to mid-funnel and convince people why they should apply to YOU, not someone else, for a credit card.

Focusing on mid-funnel users inevitably means your total SEO traffic will be lower (but that’s going to happen anyway since SGE/ChatGPT will eliminate the TOFU tire kickers that aren’t yet – or never will be – in a position to buy your product).

But the great news is, it also means your overall conversion rates will rise since the mid-funnel users who find you will already know what they need. Your entire SEO value chain will actually become more efficient as your funnel conversion rates improve.

Another mid-funnel search example.

Here’s how to win with SEO in 2024.

You have to build your SEO around who you specifically are. You have to dial everything back to your own, very specific value proposition – why you exist, why your business is better than others.

Say your value prop is insuring people who would otherwise be denied insurance because they’re classified as high-risk due to past DWIs, young age, whatever. Do your SEO around that specificity – don’t just do SEO around saving people money on car insurance (which is what everyone who sells car insurance says).

Then write compelling, MOFU content based on those specifications. That way, when someone clicks through from a gen AI search box to your website, you can convert them.

Taking all this into account, here are six key things I think you should do in 2024 to win with SEO.

 

1. Be much more deliberate in targeting the right users with the right keywords.
2. Align all of your content with your actual buyer personas.
3. Use modifiers on those head keywords you previously targeted, like “price”, “reviews” and “features.”
4. Publish content that compares you to your competitors.
5. Run individual Google searches on every topic idea on your content roadmap – if you see an SGE box show up to answer the question, consider removing the idea from your roadmap.
6. Look at “People Also Ask” (PAA) phrases for ideas on deeper questions.

Here’s what else you need to know about SEO in 2024.

1. Talk to your customer.

Really take the time to understand their buying journey. When I was at SurveyMonkey, everything I learned by directly talking to customers flew in the face of everything I’d been taught in the office. We thought, for example, that people used our surveys to collect data. But then we learned that people use surveys themselves as a KPI. That never would’ve occurred to us. But that’s why the value prop we first built didn’t resonate with a big part of our audience.

2. Be flexible.

Make plans, of course, but assume surprises will come your way. And then make plans for those, too. What will you do, for instance, if you go live with MOFU-focused SEO, and Google decides to turn off SGE for whatever reason? What will you do if Google launches gen AI, and it fails, and everyone moves to ChatGPT? Really figure out multiple plans for whatever big gen AI SEO changes might occur.

3. Build an SEO lead funnel that can survive a reporting blackout.

When the full version of Google SGE first comes out, it will probably be missing a lot of the dependencies people like you have come to count on, like Search Console and Google Analytics. That’s because it will take much longer for Google’s dev team to re-architect them to work well with SGE. So you’ll want to figure out how to understand your users, search channels and search results if, with SGE, you don’t have all the data you have today.

 4. Don’t expect accurate results from keyword tools once the full version of SGE goes live.

Keyword tools won’t work the same way – and won’t be accurate – when the full version of SGE goes live. Think about it – if the gen AI box takes up the whole fold on a Google SERP, even if you rank at number one, nobody will see it, and the click-thru rate will be totally wrong. (I’m very curious to see what happens to keyword companies.)

Two predictions for 2024 and beyond.

Lastly, Eli offered up two predictions for Google SGE you’ll want to keep in mind.

The ad market is going to change.

Google doesn’t usually take this long to reveal something. But right now, SGE is just a very small, opt-in beta available only in America, Japan and India.

I’ve heard one of the things holding Google back from a full launch is concerns around revenue – especially as we get into Q4 – because they’re trying to figure out the best way to do ads in SGE. They can either leave ads the way they are right now, which means a lot of users might not see them once gen AI is sitting on top of them. Or they can figure out how to put ads within gen AI, which I think is the more likely option.

But that’s going to be complicated. It would be a brand-new product Google will need to sell to the market. And it will change the ad market – there could be fewer impressions available, keywords could become more competitive, you might have fewer queries to bid on, etc. And that would greatly impact Google’s bread-and-butter business.

The best thing working for Google right now is that ChatGPT’s usage is plateauing. But Google isn’t dumb enough to count it out altogether – ChatGPT continues to innovate, and Google knows it’s still very much a threat to them.

There might be long periods of turmoil.

Every time Google does something big, there are always lawsuits or calls for Congress to weigh in. And SGE will be the biggest change the world has ever felt in the history of Google search, affecting roughly 80% of search results (compared to the mere 11% impacted when Google updated its algorithm in 2011 with the Google Panda update).

For every SGE winner, there’s going to be a loser. And every loser is going to want to figure out how to sue Google or go to Congress to fix it.

But none of the noise will revert things back to the way they are right now. Even if Google has to pay some fines, you’re never going to get your old search results back.

Google also always launches – then changes – things. So they’re going to launch SGE, and everyone’s going to freak out. Then they’re going to change SGE in some way, and everyone will freak out again, but for completely different reasons.

I’m already seeing it, and you probably are, too if you are in the beta (sign up here to see for yourself). You might see a generative response for a query one day; the next time you query the same term, it’s not generative. Or you receive a different gen AI response than the one you received for the exact same query made earlier in the day.

Most CMOs – and certainly CEOs – aren’t aware all of this is going to happen. So I expect a lot of panic from Fortune 500 companies that rely on content to drive search traffic when it occurs.

Remember that all of these changes won’t necessarily impact the total demand for your product category, they’ll impact how people discover your specific product within that category. Embrace these changes and build contingency plans for an uncertain future.