05 July 2023 |

AI will bring marketing fundamentals to SEO guest post by Nigel Stevens

By Tracey Wallace

👋🏻 Nigel from OGM here again with a guest spot. 

Two weeks ago, I wrote about How SEO is harder—but not as complicated—as you think

Main takeaway from that one – “yes, SEO has a whole bunch of weird complications and technical edge cases, but for most SaaS companies, it’s about doing content well, which is hard.”

This week, I’m back (yay?)—but it’s getting weirder. 

There’s no choice . . . it’s inevitable . . . we must now talk about AI.

We live in a world of extremes, and I see many opinions on the impact of AI on search gravitating to polar opposites:

  • Meh— AI search, whatever. This chat stuff is goofy, doesn’t matter. Biz as usual.
  • SKY IS FALLING SEO IS DEAD LONG LIVE CHATGPT

I don’t fall into either camp, but it’s a complex topic that can’t be fully explored in this essay. Instead, I’ll focus on the parts I personally find most compelling. 

What I cover:

  • Google’s AI search experience has a LOT of unknowns. While I believe a chat interface within search will win out in the long-term, predicting the next 1-2+ years is really tough.
  • Given all these unknowns, my team and I are focusing on existing trends that we think will maintain—or even accelerate—with AI:
    • Monetizing commoditized information gets increasingly harder over time
    • Attention is hard to capture in 2023, but SEO often doesn’t take this into account.
    • Despite all this, people still use search to solve problems and influence purchases, which comes back to the fundamental principle of connecting with users when they need to solve a problem.
  • Going forward, success in SEO and Content will be about attention and experience:
    • Focus on capturing and maintaining attention for the right audience
    • Don’t make UX and design an afterthought; it should be part of your hypothesis for doing anything
    • Every piece of content should either a) Show how you solve a user’s problem with your product, and/or; b) Blow them away with a differentiated experience that makes the touchpoint worthwhile and not a speck of dust in the wind. 

A bit of background

If you’re already caught up on Google’s SGE, skip to the next section. 

For people who haven’t been in the weeds, a bit of background:

Google has launched the beta of “Search Generative Experience” (SGE), sometimes referred to as ‘AI search’ (see their official blog posts here). It involves:

  • A ChatGPT-like feature at the top of SERPs


  • New search features to add more personal sources, such as Perspectives

So what does it all mean?

How will SGE impact SEO? A method for forming your opinion

You see a lot of hot takes on SGE, but the truth is 99% of them are making a LOT of assumptions about the future—in every direction (positive, negative, whatever).

We have to separate what we do and don’t know.

What we know: 

  • Google is testing an SGE experience. It has pros and cons, and many people have shared opinions on this. 
  • It’s showing up for lots of queries in the beta version. 
  • It currently cites 3 “sources” for many of these queries, which could end up being the new most-valuable SERP real estate.
  • Google has stated––and the UI implies––that they want to push people to engage with chat vs starting over with new queries.

That seems like a lot of change. But what do we not know?

  • How strong is the SGE search product at scale? Will it work as intended? How much of search does it roll out to—what verticals / types of queries? → We have early signals on this from the SGE beta, but it could change.
  • How does it evolve over time? We know what it looks like now, but what will it look like in 3/6/12/18/24+ months? Will it look more or less like the beta?

    Will the cost of offering SGE to users go down?

    Do the 3 ‘cited sources’ links change to more, or fewer? → It’s hard to overstate how much this experience could change as Google gathers data and tests different layouts.
  • How do users interact with–– and like––the chat interface at scale? Of all the assumptions around SGE search, this is the most uncertain.

    I personally believe that in the long-run, it will be a disturbingly good product with high adoption (just my gut feeling based on some ‘ah-ha- moments with ChatGPT).

Short and medium-term, though—I could see many people using it for many things, but as others have pointed out, a decent subset of humanity barely knows how to search effectively on Google right now!

Keep in mind that you’re hearing about the beta experience from early adopters, who are NOT like the rest of humanity (and Google is one of the rare products that is truly geared towards all of humanity)

This is NOT a prediction; rather, a reminder that with such variable outcomes for all three questions, predictions are not as easy as they appear. 

Maybe SGE gets very good very quickly, and people adopt it en masse? Maybe it takes a long time to roll out, and people adopt it slowly? Either outcome (or something in between) seems plausible to me. 

With so much uncertainty, our approach is to identify existing trends that we believe will not be slowed—or maybe even exacerbated—by AI. Our thinking: regardless of how exactly it shakes out with SGE, let’s focus on areas that will create more value in today’s world based on what we already know.


Now I’ll dive into those trends, and provide some historical context*.

*Keep in mind that this is a simplification, so chill, PhD SEO historians

Commoditized information gets eaten by tech

Google was the first search engine to effectively put a list of organic results (pages / blue links) on top of all the world’s online information. This seems obvious now, but it was a game changer back then. 

Less information on the web meant less competition. 8 and 9 figure businesses were built by just publishing information, doing weird SEO stuff to get traffic, and selling lots of ads. 

For some perspective, when you throw it on a graph, there was practically 0 data in 2000 relative to 2020 [source]:

Then around 2014, search began to get a lot smarter. The predecessor to the coming AI era was the featured snippet era:

Example: 

Fundamentally, AI chatbots at the top of search are the same general concept as featured snippets—which have been around for almost 10 years now! 

This was the first sign that Google was going to correct inefficiencies in search, and make it harder to monetize commoditized information. 

Let’s not forget that people used to make a great living from simply ranking for something like “what time does the super bowl start”—and then it went away in the blink of an eye. 

Historically, a persistent element of SEO has been finding ways to monetize commoditized information by either:

  • Capturing traffic and selling CPM ads
    • E.g., “here’s what time the super bowl starts, can I interest you in some Viagra?”
  • Upselling people to a product or service (direct or affiliate) 
    • E.g., “the super bowl is at 6:30pm, oh, while I have you, wanna buy a TV for $3,000?”

A crude summary of how this model works

Google saw this and was like, “hmm, that’s kinda silly. We can probably just figure out that the Super Bowl starts at 6:30pm and just tell people directly . . . let’s do that”. 

Which then led to:

Kinda makes more sense! There’s not a very complex need here. And if someone really wants to know other stuff about the Super Bowl or throwing a party, they can click through to a SERP result.

You might look at this as, say, a content manager at a SaaS company, and think “ok cool, but what does this have to do with me?” 

You probably don’t want to hear this, but our super bowl time example is the same underlying structure as answering “what is customer support” with “it means supporting customers, here are some generic best practices and stuff, now would you like to buy my $75K per year software product?” 

The uncomfortable truth: a decent chunk of the traffic you’re goaled against would probably go away if featured snippets were better and more thorough. 

As the quick answer improves and becomes conversational, it will weed out more and more people who don’t have a burning desire to learn deeply or solve a problem. 

Major takeaway: capturing traffic from commoditized information has been declining for years and will only worsen as ‘featured snippets’ (or soon, SGE) get much better at answering questions that don’t require deeper learning. I predict that for most SaaS companies, traffic will decline but they won’t notice much impact to their bottom line—which will shine a light on how misguided many strategies have been (some of mine included!).

Attention is getting harder to capture, but this is often ignored

The world has changed a lot since Google 1.0, yet B2B marketing hasn’t evolved at quite the same pace. 

I believe it’s a lot harder to capture attention and monetize search traffic for 2 main reasons:

  • Massive shifts in user behavior and technology
  • Search engines incentivizing copycat sameness

I’ll tackle these one at a time. 

Massive shifts in user behavior

In the Google 1.0 era, there was no YouTube, podcasts, mobile phones—none of that stuff! 

Attention was easy to come by, so you could just do weird stuff to rank and turn it into $$$.

Many of today’s execs were around during, or just after, the Google 1.0 days. I believe this era shaped today’s corporate expectations around SEO: 

It’s hard to overstate how much people’s behavior with technology has changed since Google 1.0. 

A few options for attention that weren’t around in the early days of the internet:

  • Incredibly smart people on longform podcasts chatting about every topic under the sun
  • Videos to explain anything you could possibly search for
  • Communities of like-minded people in specific niches to answer questions in an unbiased less-biased way

Not to mention social media, text message platforms, endless 8-second dance videos, etc. 

The world has evolved massively, but B2B marketing has only caught up to bits and pieces of these changes.

When you look at how much user behavior has changed, Content and SEO teams still tend to publish big meaty pieces of text content that aren’t in line with how most people consume most information in their lives. 

I know because I’ve been very guilty of this.

And that’s because the incentives of SEO have made people miss the forest for the trees.

Search engines incentivize copycat sameness , i.e. the commoditizing of the tactics.

Basic marketing rule: when everyone does the same thing, it’s a lot less likely to work.

I remember when featured snippets were showing up for more and more queries in 2015-2017. It was like shooting fish in a barrel! You could get a surge in traffic, and others hadn’t caught on yet. 

Another cheat code around the same timeframe was the early days of content optimization tools. 

If you used MarketMuse or (later) the fabulous Clearscope, it was a fast track to ranking well. So what happened? 

Everyone learned how effective featured snippet + content optimization was, and competitive segments of the web (like B2B SaaS) morphed into gigantic, over-optimized copycat clones of never-ending Q+A (shout out to the magnificent John-Henry Scherck for writing about this first ← a seriously great read).

For anyone in SEO, the game evolved to “how do we write THE MOST optimized piece of content and capture ALL of the featured snippets?” 

Fast forward a few years and plenty of smart companies have glossaries that get tons of search traffic, yet don’t even pretend to make an effort to tie it into their product or value prop. 

Everyone has access to a version of the same tool, and (almost) everyone is running a version of the same playbook. SEO got very preoccupied with optimizing for ranking, at the expense of thinking about why a potential customer would stay on the page if you happen to earn their click.

I definitely fell prey to this in the past, and started to miss the forest for the trees. 

A simple way to think about how SEO evolves:

If incentives make everything look the same . . . do ya think people won’t notice? 

It’s hard to “prove” this conclusively, but anecdotally, I hear lots of people talking about how every website has boilerplate SEO content. And it makes sense with the psychology of pattern recognition. 

Even if you genuinely have great content, if it looks like the rest, will you even get people’s attention when there are so many other options? 

SEO still works great in 2023 when executed well

Despite all this, SEO can still work wonders. I know this because I still see it firsthand in our work at OGM, and through talking to others in the industry. 

It’s pretty simple: people have problems, and if you can connect with them while they’re trying to solve their problems, you can influence a sale.

In some instances they have to wade their way through so-so content to get the solution (content can succeed in spite of itself). But I continue to see organic content show up as a strong source of discovery or influence on closed/won deals for B2B SaaS brands. 

Now, where you connect with them via search will always change. Trying to do it through commoditized information or copycat overly-optimized content was already hard and will only get harder. 

SGE will likely eat into specific page or query types that we’ll have to adjust to—Ross Hudgens has some good thoughts here

But with B2B SaaS in particular, many companies are solving very specific problems that generic answers are just not likely to resolve. 

Our bet: people will always have complex problems that need to be solved, and if you can solve their problems and make it discoverable in search, there’s no reason to think that goes away. 

SEO must shift to a laser-focus on solving problems and leaving a strong impression

In the old world of SEO, your first goal was to rank (doing whatever it took), and then you tried to “optimize for conversion”. 

In the new world of SEO, if you don’t have a strong upfront reason to think you can capture attention and get someone to do something, then don’t even bother. 

As the incomparable Ben Thompson puts it in a May update from Stratechery (emphasis mine): 

The amount of content available to end users is infinite, and covers every possible niche; AI is only going to make the abundance more overwhelming, and perfectly customized to boot.

In this world the only scarce resource is attention: even if a user is “second-screening”—on their phone while watching TV, for example—they are only ever paying attention to one piece of content at any given moment. It follows, then, that value is a function of attention, because value is always downstream from scarcity.

Our SEO hypothesis going forward is informed by this principle, and is quite simple. Content must do one of two things to justify its existence:

  • Solve a problem for the audience that shows the value of your product 

If someone gives you a reason to think they are trying to solve a problem, they have “higher intent”, and it’s easier to earn their attention

Example: Mosaic, a strategic finance platform, provides the best answer for creating a waterfall revenue forecast, and also shows how their product solves the problem even better.


  • Capture attention via a unique experience / point of view / dataset, and leave an impression for future consideration.  
    → If someone does not give you a reason to think they are urgently trying to solve a problem, they have “lower intent”, and you have to blow them away to earn their attention and be memorable

    Example from a while ago: Intercom published books when others weren’t (shout out to the great John Collins). They had
  • Unique opinions – their own strong opinions.
  • Unique data – actual examples from Intercom.
  • Unique presentation – actual book, with a compelling distribution.

This sounds obvious, but in my estimation, the majority of B2B content lives in the void between these criteria: it either doesn’t SHOW how to solve a problem with your product, or doesn’t have sizzling data or experiences that resonate with people. 

Think of how many individual pieces of media you interact with on a daily basis. How many actually leave an impression and trigger you to take an action in the future? 

Probably not many—and you want your marketing touchpoints to be in that minority. 

Focus on UX and content design to keep users engaged

These are platitudes at this point —“user experience is king, blah blah blah”—but the truth is, it’s hard and most websites don’t do a good job. 

A simple equation – when you combine 

  • a premium on attention, with;
  • the trends of evolving user behavior and copycat content blindness, it means; 
  • a UX focus to distinguish your content will go from nice-to-have to table stakes. 

Tracey has shared some excellent resources on this in the past; here are a couple examples:

A few principles we are building into our playbook going forward, most of which Tracey covers above:

  • Make it easy for users to jump to various parts of your content as they try to solve their problems
  • Provide people with multiple formats to consume your content, including video, audio, and summaries
  • Make it clear what value you offer and get to the point, or else why wouldn’t someone go somewhere else? 
  • Signal through design and UX that you don’t have another piece of copycat content, but an experience worth staying for. Do something different.

In closing

If it makes you feel any better, I’m a pretty experienced SEO professional and went through a stage of existential bUt wHaT iS SEO aNyMorE MaaAaAaNnNN earlier this year. 

I’m fortunate to know some of the best SEO people in the world, and across the board everyone has shared the same sentiment at some point. 

Point being: it took me a good while—and LOTS of pacing around my office—to reach the conclusions I shared in this essay, so I don’t blame anyone for being unsure or asking tough questions. 

I don’t pretend to have the answers, and you shouldn’t blindly listen to me. 

To sum up this whole essay: we have to ‘unlearn’ a lot from the prior eras of SEO — particularly that traffic is a) created equal, and; b) the main goal. It will likely turn out that everyone’s traffic was inflated by inefficiencies in search—and that’s OK.

Instead, focus on the fundamentals of showing users how you can solve their problems and leaving them with a positive impression, and good things should happen. 

It’s more complicated than that . . . but also, is it really?

More thoughts and predictions from smart people

These are some of the smartest people in the game who have all played a part in informing the views I shared in this essay: