The Marketer’s Playbook for the Intelligent Web

The Marketer’s Playbook for the Intelligent Web ![rw-book-cover](https://img.youtube.com/vi/w0oMGzkEcao/maxresdefault.jpg) 🔗 **Source:** [The Marketer’s Playbook for the Intelligent Web](https://www.youtube.com/watch?v=w0oMGzkEcao) > [!note]+ Document Notes > tags: #ai-search #seo #content-marketing #website-optimization #digital-marketing #attribution > author: WP Engine > source: YouTube > reference: https://www.youtube.com/watch?v=w0oMGzkEcao > date created: 2024-12-19 14:30 > > # 🏷️ Title: The Marketer's Playbook for the Intelligent Web >  ## 🧰 Action Inbox - [H1] Audit website from fresh perspective — evaluate if users understand what you do and why you matter #content-marketing - [H2] Implement flexible architecture with structured content — enable composable/headless setup for AI readability #website-optimization - [H3] Build new attribution framework combining traditional metrics with AI visibility — track mentions, citations, sentiment alongside conversions #digital-marketing ## 🧭 Table of Contents - [[#📝 Summary]] - [[#📖 Main Content]] - [[#🗒️ Notes]] - [[#📚 References]] - [[#🗣️ Paraphrase]] - [[#📰 Newsletter]] - [[#✅ Action Items]] - [[#🧪 Hypotheses]] - [[#📊 Methodology]] - [[#❓ Questions]] - [[#🔮 Future Research]] - [[#💡 Implications]] - [[#🧠 Learning Styles Assessment]] - [[#📊 Accessibility Assessment]] - [[#🌍 DEIB Assessment]] - [[#✅ Logical Validity]] - [[#🛡️ DARVO Analysis]] - [[#🎯 Conclusion]] ### 🔗 References - [Don't Make Me Think (Book)](https://www.amazon.com/Dont-Make-Think-Revisited-Usability/dp/0321965515) - [WP Engine AI Toolkit](https://wpengine.com/) ## 📝 Summary - Over 50% of searches now end without clicks, accelerating dramatically in the last year - Traditional attribution is breaking down as more traffic becomes "direct" or unattributable - Websites need rearchitecting around flexible foundations, structured content, and fast delivery - Success requires combining traditional metrics with new AI visibility measurements - User experience remains the foundation—what works for humans works for AI ## 📖 Main Content The session addresses how marketers must adapt to AI-driven search where zero-click searches dominate and traditional attribution models fail. Rebecca Baron (WP Engine), Darcy Kurtz (CMO at WP Engine), and James Baldakino (Head of SEO at Ellipsis) discuss practical strategies for website optimization and measurement in this new landscape. Key challenges include declining SEO traffic that's shifting to direct/referral channels, faster conversion cycles from pre-qualified AI traffic, and the need to serve both human users and AI bots effectively. The solution involves three pillars: flexible/composable architecture, structured content with proper metadata, and scalable technical infrastructure. The speakers emphasize that this evolution builds on existing SEO principles rather than replacing them entirely. Success still depends on answering user questions effectively and providing excellent user experience, while adding new layers of technical optimization for AI consumption. ## 🗒️ Notes #### ⏱️ [00:00](https://youtu.be/w0oMGzkEcao?t=0) — [H1] The Scale of AI Search Impact - Key idea: More than 50% of searches now end without clicks, accelerating wildly in the past year - Evidence: "More than 50% of searches are now ending without a click...it certainly has accelerated pretty wildly over the last year alone" #### ⏱️ [03:45](https://youtu.be/w0oMGzkEcao?t=225) — [H2] The Magic Wand Misconception - Key idea: Marketers wrongly believe there's a quick fix for AI search optimization - Evidence: "The biggest misconception is...that there's a magic wand you can wave and you can just win overnight" #### ⏱️ [05:12](https://youtu.be/w0oMGzkEcao?t=312) — [H1] Everyone is Learning Together - Key idea: No one has AI search figured out yet—teams should start experimenting rather than waiting - Evidence: "No one's got this figured out yet...the advantage right now is really going to the teams who just get start" #### ⏱️ [07:30](https://youtu.be/w0oMGzkEcao?t=450) — [H1] Traffic Attribution is Breaking Down - Key idea: SEO traffic is declining while direct/referral traffic increases, making attribution fuzzy - Evidence: "We are seeing SEO traffic go down. Traffic is shifting into like direct or referral traffic at best" #### ⏱️ [08:45](https://youtu.be/w0oMGzkEcao?t=525) — [H2] Higher Quality Traffic from AI - Key idea: AI-referred visitors convert faster and at higher rates because they're pre-qualified - Evidence: "When they get to your site they are converting much much much faster and at a higher rate...because they've already learned in the LLMs who you are" #### ⏱️ [12:15](https://youtu.be/w0oMGzkEcao?t=735) — [H2] Content Must Answer Questions Directly - Key idea: Users expect immediate answers without scrolling through preamble or fluff - Evidence: "We are living in a world where people ask questions and expect answers even faster than they did before" #### ⏱️ [15:30](https://youtu.be/w0oMGzkEcao?t=930) — [H3] Conflicting Content Length Advice - Key idea: Expert recommendations on content length are contradictory, changing within weeks - Evidence: "Few weeks ago...content has now extended dramatically...then last week, I saw the complete opposite" #### ⏱️ [17:45](https://youtu.be/w0oMGzkEcao?t=1065) — [H1] When to Rebuild vs Rearchitect - Key idea: Most companies need rearchitecting rather than rebuilding to organize around topics and expertise - Evidence: "I don't think it's really for most companies it's not a rebuild it's probably a rearchitecture" #### ⏱️ [20:30](https://youtu.be/w0oMGzkEcao?t=1230) — [H2] Websites as Distribution Systems - Key idea: Websites must serve both humans and machines as systems for distributing expertise - Evidence: "Thinking of it as like it's a system for distributing who you are and your expertise...it has to serve both humans and machines" #### ⏱️ [23:15](https://youtu.be/w0oMGzkEcao?t=1395) — [H2] Start with Human Usability - Key idea: AI optimization should begin with ensuring the site works well for real people - Evidence: "You have to start from there because ultimately customer experience and user experience is the core of everything" #### ⏱️ [26:00](https://youtu.be/w0oMGzkEcao?t=1560) — [H2] Three Pillars of AI-Ready Architecture - Key idea: Flexible foundation, structured content, and fast scalable delivery are essential - Evidence: "An AI ready website...comes down to three things...flexible foundation...structured content...fast and scalable delivery" #### ⏱️ [28:30](https://youtu.be/w0oMGzkEcao?t=1710) — [H3] Vector Databases Enable Semantic Matching - Key idea: Vector databases allow content to appear for related prompts without exact keyword matches - Evidence: "It allows your content to show up for prompts that are related to the content without having the exact keywords" #### ⏱️ [32:45](https://youtu.be/w0oMGzkEcao?t=1965) — [H1] Attribution Addiction Must End - Key idea: Marketers are addicted to direct attribution but must wean off this "drug" as it becomes impossible - Evidence: "We are all addicted to this drug called attribution...It's not possible anymore. We all have to wean ourselves off the drug" #### ⏱️ [36:20](https://youtu.be/w0oMGzkEcao?t=2180) — [H2] Building a Mosaic of Metrics - Key idea: Success requires combining traditional metrics with AI visibility indicators to paint a complete picture - Evidence: "You've got to create this sort of mosaic that shows...this is how this is all starting to come together" #### ⏱️ [40:15](https://youtu.be/w0oMGzkEcao?t=2415) — [H1] Pressure Test Your Value Proposition - Key idea: If AI had to explain what you do and why you matter, would it give the right answer? - Evidence: "If AI had to explain what we do, why we matter, what would it say? And is it the right thing?" ## 📚 References - [Don't Make Me Think by Steve Krug](https://www.amazon.com/Dont-Make-Think-Revisited-Usability/dp/0321965515) - [WP Engine AI Toolkit](https://wpengine.com/) - [Advanced Custom Fields (ACF)](https://www.advancedcustomfields.com/) ## 🗣️ Paraphrase - Traditional search attribution is becoming unreliable as AI systems increasingly answer questions without sending users to websites - Successful websites need flexible architectures that can distribute content across multiple channels while maintaining structured data - Measurement strategies must evolve from direct click attribution to broader brand visibility and engagement metrics across AI platforms ## 📰 Newsletter **AI Search is Reshaping WordPress Strategy** The data is in: over half of searches now end without clicks, fundamentally changing how WordPress sites need to be built and measured. WP Engine's latest insights reveal that while traditional SEO traffic declines, AI-referred visitors convert faster because they arrive pre-qualified. **Key takeaways for WordPress professionals:** - **Architecture matters**: Headless/composable setups with ACF for structured content - **Speed is critical**: AI bots are hitting sites harder than ever - **Attribution is evolving**: Combine traditional metrics with AI mention tracking The good news? WordPress's flexible content management capabilities position it well for this shift. Focus on answering questions directly, structuring content for both humans and machines, and building sites that load fast under bot traffic. Resources: [Learn WordPress](https://learn.wordpress.org/), [Developer Docs](https://developer.wordpress.org/), [WP-CLI](https://wp-cli.org/) ## ✅ Action Items #### WordPress - Core: Investigate AI-friendly structured data standards for posts and pages #ai-search - Core Performance: Prioritize bot traffic optimization in performance budgets #website-optimization - Documentation: Create AI search optimization guidelines for developers #seo - Community: Develop meetup content on composable WordPress architectures #content-marketing #### OSPO - Strategic Alignment: Track AI search visibility metrics in quarterly reviews #digital-marketing - Developer Support: Provide training on headless WordPress for AI optimization #website-optimization #### Community - Contributors: Build plugins for semantic content structuring and vector database integration #ai-search - Maintainers: Establish best practices for AI-readable content organization #content-marketing ## 🧪 Hypotheses - Websites with composable architectures will see better AI search visibility than monolithic sites - Structured content using custom fields correlates with higher AI citation rates - Sites optimized for bot traffic will handle AI search indexing more effectively - Direct traffic conversion rates will increase as AI pre-qualifies visitors ### 📊 Methodology - A/B test structured vs unstructured content for AI citation rates - Measure correlation between site speed and AI bot crawl success - Track attribution patterns before and after architecture changes - Survey conversion rates by traffic source (traditional vs AI-referred) ### ❓ Questions - How can WordPress core better support AI-readable structured content? - What specific technical optimizations most improve AI search visibility? - How should attribution models evolve to capture AI-driven conversions? ### 🔮 Future Research - Integration patterns between WordPress and vector databases - Impact of different headless architectures on AI search performance - Evolution of content strategy for AI-first discovery ### 💡 Implications - Practice: Content teams need training on AI-optimized writing and structuring - Policy: Marketing measurement frameworks require fundamental restructuring - Security: Increased bot traffic demands enhanced DDoS protection strategies - Education: WordPress developers need composable architecture skills - Funding: Marketing budgets must account for longer attribution cycles ### 🧠 Learning Styles Assessment Reading level estimate: 12th grade (business/technical content) #### 👀 Visual learners - Present: Conceptual framework diagrams for three-pillar architecture - Improve: Traffic attribution flow charts, before/after metrics visualizations #### 🔊 Aural learners - Present: Full video discussion with expert dialogue - Improve: Audio summary of key action items, recorded implementation guides #### 🗣️ Verbal learners - Present: Clear definitions of technical terms, business analogies - Improve: Glossary of AI search terminology, case study narratives #### 🧩 Kinesthetic learners - Present: Actionable website audit checklists - Improve: Hands-on WordPress configuration tutorials, implementation labs #### 🧠 Logical learners - Present: Structured three-pillar framework, metric correlation analysis - Improve: Decision trees for architecture choices, ROI calculation models #### 🧑‍🤝‍🧑 Social learners - Present: Panel discussion format with expert dialogue - Improve: Community forums for sharing AI search experiences, peer review sessions #### 🧘‍♀️ Solitary learners - Present: Self-assessment questions for website readiness - Improve: Individual reflection worksheets, personal progress tracking tools #### 🌿 Naturalistic learners - Present: Evolution analogies for AI search adaptation - Improve: Ecosystem thinking about content distribution, sustainable optimization strategies ### 📊 Accessibility Assessment #### Screen reader semantics - Present: Video has clear speakers and topic transitions - Improve: Transcript with proper heading structure, speaker identification #### Keyboard navigation - Present: Standard YouTube controls work with keyboard - Improve: Timestamped navigation links, chapter markers #### Contrast and readability - Present: Video quality allows reading of presentation slides - Improve: High-contrast slide templates, larger text overlays #### Media and captions - Present: Full transcript provided - Improve: Professional captions with technical term spelling, downloadable transcript ### 🌍 DEIB Assessment #### Cultural breadth - Present: Business perspectives from agency and in-house marketers - Improve: Global examples of AI search impact, non-US case studies #### Inclusive language - Present: Professional terminology with explanations - Improve: Plain-language alternatives for technical jargon, avoid industry assumptions #### Neurodiversity support - Present: Clear section breaks with topic transitions - Improve: Visual agenda markers, summary bullets for key points #### Representation - Present: Mixed gender panel with diverse professional backgrounds - Improve: Broader range of company sizes and industries represented ### ✅ Logical Validity #### Strong reasoning - Deductive strength: Three-pillar architecture framework logically follows from AI requirements - Inductive strength: Traffic pattern observations support broader attribution changes - Causal reasoning: Clear connection between AI search growth and attribution challenges - Analogical strength: Billboard attribution analogy effectively illustrates measurement challenges #### Fallacies - Appeal to novelty: Some emphasis on "new" without acknowledging continuity with existing practices - False dilemma: Slight implication that traditional and AI search are mutually exclusive rather than coexisting #### Tone of voice - Overall tone: Professional, collaborative, and pragmatic with appropriate technical depth ### 🛡️ DARVO Analysis - Deny: No concerning patterns detected - Attack: No attacks on accusers present - Reverse victim/offender: No victim/offender dynamics present - Emotional cues and rhetoric: Professional discussion tone throughout ## 🎯 Conclusion The shift to AI-driven search represents an evolution rather than revolution in web optimization. Success requires rearchitecting websites around flexible foundations, structured content, and enhanced user experience while developing new measurement approaches that combine traditional metrics with AI visibility indicators. WordPress sites are well-positioned for this transition given their content management flexibility, but organizations must start experimenting now rather than waiting for perfect solutions. ## 🧱 Refactor Appendix ### 📚 Literature Split Pad Purpose: Stage CEE units that become Literature notes. Split level: H3 ### [H1] AI search visibility requires flexible content architecture > [!quote] [The Marketer's Playbook for the Intelligent Web](https://youtu.be/w0oMGzkE --- ## Full Transcript [00:17] Hi everyone. Thank you so much for [00:19] tuning in to today's session all about [00:21] creating a marketer's playbook for the [00:24] intelligent web that we're dealing with [00:26] today. More than 50% of searches are now [00:30] ending without a click. And even though [00:33] zero-click search is not a new concept [00:35] to a lot of us who have been around SEO [00:37] for a long time, it certainly has [00:40] accelerated pretty wildly over the last [00:42] year alone. So today, we're really going [00:44] to dive into how do you structure your [00:47] content and your website's technical [00:50] foundation to better prepare for this [00:52] AIdriven world we are now in the throws [00:55] of. Um, and we'll also really dig into [00:58] the metrics. How are they evolving [01:00] today? Um, especially as marketers, [01:03] we're dealing with these visibility [01:05] metrics and how do we translate those [01:07] into a narrative for our CEOs and CFOs [01:11] and boards to truly understand so that [01:14] they can ultimately hopefully support [01:16] your next big marketing bet. I'm Rebecca [01:19] Baron. I am the head of organic growth [01:20] here at WP Engine and I will be [01:22] moderating today's discussion between WP [01:25] Engine's own chief marketing officer [01:27] Darcy Kurtz and James Baldakino who is [01:30] the head of SEO at the digital agency [01:32] Ellipsis. So thank you both so much for [01:34] joining me today. I'm I'm really excited [01:36] to dive in. [01:38] >> Thanks for having us. [01:39] >> Yeah. [01:40] >> Great. I want to kick us off with a [01:42] prompt for humans, not for bots. Um, in [01:45] one sentence, how would you uh how would [01:49] you describe the biggest misconception [01:50] or AI myth that marketers still have [01:53] today about AI search? James, I'd love [01:56] for you to kick this one off. [01:58] >> The biggest misconception is this [02:00] similar to the misconception we've [02:02] always had with SEO in general, which is [02:03] that there's a magic wand. You can wave [02:06] and you can just win overnight. being [02:11] found in AI search is let's call it a [02:14] version 2.0 though of uh being found in [02:18] search engine, it's an evolution, not a [02:21] completely new methodology. Um so, so as [02:27] old, let's call it old SEO, it's not [02:29] really old, but old SEO used to be, [02:32] there's no magic wand. There's no do [02:34] these three things right and you will be [02:36] at the number one. Anyone saying that is [02:40] probably selling you stuff you shouldn't [02:42] be buying. [02:45] It's a great point, Darcy. How about [02:47] you? [02:48] >> Yeah, I mean, I totally agree. Um, it's [02:51] it's funny. I I think there's there's a [02:53] lot of fear baked into AI conversations [02:56] right now. It's like everyone else has [02:58] it figured out and like we're bombarded [03:00] in our inbox with newsletter after [03:03] newsletter of like, we we know how to do [03:04] it and you're behind and you should do [03:06] it. Like the bottom line is like no [03:09] one's got this figured out yet. This is [03:10] still new. like with SEO, like did we [03:13] ever fully figure it out? To your point, [03:15] James, right? Like there's always [03:16] something that's changing. So, for me, [03:19] it's really like uh how do we how do we [03:22] break this thing that's out there in the [03:24] ether that we're all behind and late and [03:26] we got to catch up? Like, we're all [03:28] right where we're supposed to be right [03:29] now. We're all learning. We're all [03:30] trying to to see where this is. And I I [03:33] think the you know the advantage right [03:35] now is really going to the teams who [03:37] just get start just start start learning [03:39] start trying some things you know um if [03:42] you're waiting for the perfect playbook [03:44] then you are late and you aren't you [03:46] aren't going to be there because there [03:47] is no perfect playbook and it's just [03:48] like jump in get started you're not [03:50] behind we're all figuring this out [03:52] together as we go. [03:53] >> I love that. That's super helpful. Yeah. [03:55] today thinking about you're creating [03:57] your own playbook and just getting [03:59] started um and not thinking that it's [04:01] one-sizefits-all [04:02] that's really really helpful. So we are [04:05] obviously well beyond this being a [04:07] theoretical problem and we are all today [04:12] seeing the AI shift show up in our data [04:15] can you know Darcy I'd love to start [04:17] with you how are you seeing it [04:19] especially from the perspective of WP [04:20] Engine show up in the data today how [04:22] this shift is happening [04:24] >> yeah absolutely we we are seeing it you [04:26] know a year ago when we were talking [04:27] about this it was like what the theory [04:30] was in the world of what was going to [04:31] happen and now we're actually seeing it [04:33] come to life in our own data. We are [04:35] seeing SEO traffic go down. Um traffic [04:38] is shifting into like direct or referral [04:41] traffic at best if you see that at all. [04:43] Um so so we are seeing that trend of [04:46] unattributed traffic uh growing. We're [04:49] also really starting to see one of the [04:51] the theories that was out there that I'm [04:53] actually excited was true is that and [04:55] when they get to your site they are [04:57] converting much much much faster and at [04:59] a high you know at a higher rate and at [05:01] a faster pace. So sales cycles are [05:02] shorter and and conversion rates are [05:04] higher because they've already learned [05:07] um in the LLMs, you know, who you are [05:09] and already been told they should buy [05:11] you. So when they get to your site, uh [05:13] they're actually expediting through uh [05:15] which is something we've really leaned [05:16] into and said, "Okay, how do we have buy [05:19] paths on our site that are faster for [05:21] these people that are are aren't just [05:23] here to learn, they've already learned [05:24] somewhere else?" Um and then, you know, [05:27] we we we see it in our our marketing [05:29] attribution. and it's just getting [05:30] harder and harder and harder, right? [05:32] Which we knew was going to happen. Um, [05:33] and the more people that start in LLMs, [05:35] attribution is just be going to become [05:37] much much fuzzier. Uh, so we're we're [05:39] seeing that in our own data now is like [05:41] our ability to really tell where did [05:43] everything come from and what's driving [05:45] conversion um is is just getting clouded [05:48] and we're having to adjust to that and [05:50] figure out okay, how do we triangulate [05:52] lots of different pieces of data to [05:54] figure out what we want to do and and [05:56] how to address it. [05:57] >> Absolutely. James, from the agency side [05:59] of things, you know, how is this showing [06:01] up for your clients and how are you [06:02] handling that? [06:03] >> I mean, listening to Darcy is like [06:05] listening to a breath of fresh air [06:06] because that is literally what I tell [06:08] all our all our clients. Uh, [06:11] unfortunately, from the agency [06:12] perspective, it sounds like I'm trying [06:14] to make excuses, but that what Darcy [06:17] said is the perfect truth in the sense [06:19] of attribution has become much more [06:22] difficult. We are we we we we've been [06:24] we've been blessed so to speak that in [06:27] the past years attribution was [06:28] relatively easy in the sense we are so [06:31] used to seeing customer journeys in a [06:34] kind of linear fashion that now that [06:37] they're broken we're we're we're we're [06:39] kind of we can't believe the truth that [06:42] that things are shifting. So yes Darcy I [06:45] could not I could not agree with you [06:47] more and I'm really glad you said that. [06:49] Uh, [06:51] and as Darcy said, and we'll get to this [06:53] probably in the in the tracking section [06:55] a little bit more. As Darcy said, you [06:57] need to build a picture which is more [06:58] around the whole picture rather than [07:02] simply looking obsessing on clicks or on [07:06] conversions from the blog or whatever it [07:08] is that you used to obsess on. that that [07:11] ship has sailed. And uh yes, zero click [07:14] has always been around, but zero click [07:17] is now significantly bigger. Uh but that [07:20] doesn't mean you're invisible. That [07:22] doesn't mean you're you're not showing [07:24] up. As always, the most basic rule SEO [07:28] has always had is does your site have [07:33] the content to answer your client's [07:35] questions properly? [07:38] And does your site, as Darcy said, guide [07:41] you to getting to where you want easily? [07:44] It is useless having the best content in [07:46] the world if your site is unusable. And [07:49] it is useless having the nicest designed [07:52] site in the world if you don't have the [07:53] content to support it. [07:55] >> Absolutely. I want to click a little bit [07:57] deeper into that cuz you you kind of uh [07:59] led to the next topic I really want to [08:01] cover with both of you which is, you [08:03] know, why is just plain old good content [08:06] not enough anymore and and what has to [08:09] be around that? How do you make good [08:10] content great content and how do you [08:13] better support it to get it found? [08:15] >> I am a strong believer that uh as I've [08:19] said earlier that it's not it's not [08:22] about for example just doing structured [08:24] data. Yes, structured data is important. [08:26] Yes, getting the technical SEO basics [08:29] are important. But the most critical [08:32] part is is uh getting the content right [08:37] and giving the right answers, avoiding [08:40] fluff and just giving we are living in a [08:43] world where people ask questions and and [08:46] expect answers even faster than they did [08:48] before. If you think about it now, it [08:51] feels a bit weird that up to two years [08:54] ago, we used to troll through pages for [08:57] the answer. It it already feels weird. [09:00] So, if your content isn't giving the [09:02] answers and is giving the the the the [09:04] preamble, [09:06] that's already uh uh a problem. So the [09:10] way I'm answering the question on why [09:12] it's not why good content alone is not [09:15] cutting is is [09:17] the the the way you measure what good [09:21] content is has also changed [09:22] significantly. [09:24] >> Yeah, it's interesting. Um, I I think [09:27] you're so right on like focus on [09:29] creating great content. Like, and you [09:31] know, few weeks ago, maybe a month ago [09:33] or so, I I was listening to some people [09:35] talk about how your content needs to [09:37] change and they're like content now has [09:40] gone the the the length of content has [09:42] now extended dramatically so that you're [09:44] you're hitting more of the long tail and [09:46] things like that. And then last week, I [09:48] saw the complete opposite, which is, you [09:50] know, GEO is now looking for short, [09:52] concise, answer first kind of stuff. [09:53] This is where things are really [09:55] changing. And so like just focus on [09:57] building great content, you know, create [09:59] great content that answers questions um [10:03] that that is structured in a way that [10:04] when AI sees it, it knows what you do, [10:07] who you are, you know, how you relate to [10:09] a broader topic, but you know, good [10:12] content is still going to be the core [10:14] and trying to just figure out all these [10:16] little optimizations like it's literally [10:18] changing in a four-week period. I went [10:20] from the polar opposite ends of how we [10:22] should be structuring it. So, I think [10:23] it's really more about um you know, not [10:26] about how do we create more content, but [10:28] how do we create content that's [10:30] understood, it's trusted, it's it's [10:33] valuable to humans because if it's [10:35] valuable to humans and structured so [10:37] that machines can interpret it, then [10:40] you're good. I think that's so helpful. [10:42] And I want to take a a further step back [10:45] because having great content is so [10:48] important, but we obviously all need to [10:50] have a place to put it. So, I think a [10:53] big question on everyone's minds with [10:55] this shakeup is a common question that [10:57] I'm sure in the agency world James [10:58] you're getting with clients and Darcy [11:00] I'm sure is no strange question to you [11:03] um going from company to company when [11:05] you look at a website when do you know [11:07] it's time to rebuild it um when it comes [11:10] to this AI world of search I is [11:13] rebuilding an important thing to ask [11:15] yourself now and if so what do you [11:18] usually expect the impact to be? Yeah, [11:20] you know, a year ago, I I actually had [11:22] this very conversation with some of my [11:24] CMO colleagues because, you know, being [11:26] in the website business, they're like, [11:27] "What are you seeing? What do you think? [11:28] What do we need to do?" And um a lot of [11:30] my conversation was, "It's not yet." [11:32] Like, we didn't know enough yet. We [11:34] weren't sure how things what did need to [11:36] change. And so, like you if even if you [11:38] did do something, the likelihood you [11:40] were going to have to do it again. So, [11:41] it was kind of in a wait and see mode. I [11:43] think we're in a very different place [11:44] today. I do think we do understand [11:47] enough about how this is working to [11:49] really know uh at least to start because [11:52] and I don't think it's really for most [11:54] companies it's not a rebuild it's [11:56] probably a rearchitecture [11:58] um to to help you um and you know for [12:01] you to know the things that I started [12:03] looking for on our site are things like [12:05] you know is your traffic flat or down [12:07] but the industry demand isn't right this [12:10] is that zero you're getting that zero [12:11] click phenomenon you know you're [12:13] creating a lot of content but it's not [12:15] showing up or it's not getting cited. Uh [12:18] is your site or still organized around [12:20] pages versus topics and expertise. Um [12:23] you know it it's just like you know can [12:26] you move fast? Are you able to make [12:28] changes really really quickly? So that [12:30] you know if if you're saying you know [12:32] I'm having some of these issues those [12:34] are signs that it's like probably time [12:36] to start rearchitecting to be ready for [12:39] this new era. um you know and the impact [12:41] of that is to me I think is pretty [12:43] meaningful as far as you know you're [12:45] going to show up more you're going to [12:46] get more of that direct traffic you'll [12:48] get your traffic back but it's going to [12:49] come through different places um you [12:52] know and I think this is really the the [12:55] biggest shift here is not just thinking [12:57] of our website as just a collection of [12:59] pages where people come to learn but [13:01] really thinking of it as like it's a [13:03] system for distributing who you are and [13:05] your expertise and like it has to serve [13:08] both humans and and machines in this new [13:10] world. And so what are some of those [13:12] steps you can start to take to [13:14] rearchitect to allow the the machines um [13:18] as well as the humans to really uh [13:20] access your data and your expertise. So [13:23] uh I don't see this as much as a design [13:25] exercise. It's really about making your [13:27] website more usable for this new [13:29] generation of things, these bots that [13:31] are coming in and using your site. I am [13:34] a very strong believer that ultimately [13:37] uh it all boils down to your website [13:40] being usable by real people. You have to [13:44] start from there [13:46] because ultimately customer experience [13:48] and user experience is the core of [13:50] everything. Once again if you have a [13:52] website which is hyper you've found the [13:55] magic formula which hyper optimizes you [13:57] for SEO that doesn't mean that your site [14:00] is actually usable. So I I this is why [14:04] I'm agreeing with you Darcy because I I [14:07] if you start from okay is my website [14:10] actually saying what I want my clients [14:12] to know why why my product is good why [14:16] they should trust me why they should [14:18] come to me can they find stuff can they [14:20] navigate it well can they go through to [14:22] can they connect one thing to the other [14:24] if they're looking for this particular [14:26] feature can they can they really have a [14:29] cohesive journey throughout the website [14:32] around it. If you do that, [14:35] good things will follow. Why? Because [14:38] it's the the answer is in is is in the [14:41] title. It's AI, artificial intelligence. [14:44] What it's trying to be is as intelligent [14:49] as a real human, arguably more, but the [14:52] point is it's trying to emulate how we [14:56] think and how we connect things. So if [14:58] you get it right for humans, you're very [15:01] much on the way there to get it right [15:03] for for the bots as well. [15:05] >> This is super helpful. So I feel like [15:07] the theme here being it's about when [15:10] your site you can tell it's ready. It [15:12] needs something. It needs a lift of some [15:14] kind to be a lot more ready. Um and then [15:17] as as James is saying it's then taking [15:20] it that step next step and when you're [15:22] doing it ensuring that everything is [15:24] connecting correctly um for that human [15:26] journey because ultimately that's how [15:28] these bots are are really moving through [15:30] websites. So I want to go into like more [15:33] actionoriented then. So say a company [15:37] they know they need the refresh of some [15:39] kind whether it's rearchitecting whether [15:41] it's updating the journey making content [15:43] more readable and findable. Um what you [15:47] know James I'd love to start with you [15:48] especially from the agency perspective. [15:51] What do you find are some of the most [15:52] impactful uh refresh activities that you [15:56] recommend for websites that'll have the [15:58] best return for them in the long run? [16:00] It's a great question. Um, in my view, [16:03] one of the most important things you can [16:05] do is genuinely audit your site from a [16:10] fresh mind perspective. Now, as a as a [16:14] site owner or as a site marketer, as a [16:16] CMO, that is really difficult, [16:18] especially if you've worked on the same [16:19] site for a very long time. Um but one if [16:23] you can offer if you can look at it from [16:25] a fresh point of view you can ask the [16:28] basic questions. The basic questions [16:30] being am I am I pointing people in the [16:34] right direction. Am I actually do people [16:37] know that I have this amazing feature [16:39] which nobody else does? Do people know [16:41] that my product is better for XY Z [16:44] reasons? Can they actually find this [16:46] information or do they have to go dig in [16:48] in the inside the the documentation [16:51] somewhere and it's in a half a sentence [16:53] somewhere down the bin the middle of the [16:55] page somewhere? [16:57] If you manage to ask those questions [17:00] neutally and you get answers neutally [17:03] then you can have a very honest take on [17:05] how you should approach the refresh. The [17:08] the refresh should start from the basics [17:10] and the basics being do people really [17:13] get it? Do people who who see our site [17:16] really understand it? What I've seen [17:18] more time and time again is that over [17:20] time a website evolves evolves evolves [17:24] evolves and at some point it becomes a [17:25] bit of a pile. Uh so my view of the best [17:30] refresh activity you can do is park [17:33] everything you know about who requested [17:35] what and where and why why that page [17:37] goes and just look at it from a neutral [17:39] perspective and say how does this map to [17:42] how I want people to discover my [17:44] product. [17:45] >> Yeah I I I like that. I think there's a [17:48] lot that can be done just about cleaning [17:49] up what you have already and and making [17:52] it more usable. Um, I'm also finding [17:56] some things on the back end that that [17:57] need to be uh rearchitected as well. Um, [18:01] and I think a as a as a marketer and not [18:04] a technician, um, this is where a lot of [18:07] marketers definitely are a little [18:09] intimidated because that it does start [18:12] to sound very technical when you start [18:13] talking about some of these things. So, [18:15] um, but when you when you kind of strip [18:17] it all down, the an AI ready website to [18:20] me is really comes down to to three [18:22] things. And the first is having a very [18:24] flexible foundation. Um this is what [18:27] people mean when they talk about like a [18:28] composable architecture or a headless [18:30] architecture. This ability to um ensure [18:34] your content isn't locked into one rigid [18:37] system. You can reuse it. You can [18:38] restructure it. You can deliver it [18:39] wherever it needs to go whether it's on [18:41] your website or an app or wherever. So [18:44] something that allows for much more of a [18:46] composable flexible architecture I think [18:49] is going to be important. Um, and then [18:52] we've talked a lot about structured [18:53] content. Um, that that's the second one, [18:56] which is how do we structure our content [18:58] that has context for it. So, not just [19:02] pages and blogs, but content that [19:04] clearly explains what is the content in [19:06] this section, how does it relate to [19:08] other things? Why does it matter? Why [19:10] would you want it? Um and so from a from [19:13] a architecture standpoint, tools that [19:16] allow you to add that structure, you [19:17] know, we we have ACF which allows you, [19:20] you know, custom fields or metadata uh [19:22] that makes your content understandable [19:23] to machines is is really valuable here. [19:26] Um and things like uh our vector [19:29] database that we have in our AI toolkit. [19:31] Um that that includes semantic [19:33] embeddings, which is one of those really [19:35] technical terms that I had to ask like [19:37] three time help me understand what what [19:38] this vector database does. And really [19:40] all it all it means is that it allows [19:42] your content to show up for prompts that [19:45] are related to the content without [19:47] having the exact keywords of your [19:49] content. So it allows your content to [19:51] show up in a lot more places than maybe [19:54] in a traditional SEO world where you [19:55] needed a word for word and it needed to [19:57] be in your content. A vector database [19:59] really allows you to show up for things [20:02] that aren't um word for word exactly the [20:04] same. So, uh, it's a really cool new [20:06] technology that that, um, enables LLMs. [20:10] And then the third thing is really just, [20:12] you know, how do we have a really fast [20:14] and scalable delivery later layer. So, [20:17] this is where, you know, James has been [20:18] talking about a lot that SEO and GEO are [20:21] not that far apart. You know, you still [20:23] have to have pages that load quickly. [20:25] You have to have pages that are able to [20:27] handle a lot of requests. Um, because [20:29] now you've got all these bots coming in. [20:31] your direct traffic is probably [20:32] skyrocketed because the the bots are [20:35] coming in at at a record rate. So really [20:37] checking all those technical details on [20:39] the back end. Um so that there's you [20:41] know you can reliably serve up your [20:43] content in a way that's really easy to [20:45] consume. So you know flexibility, [20:48] structure and and scalable delivery you [20:51] know your technical backend. Those to me [20:53] are the really the the places to to [20:55] focus on the backend part to make sure [20:57] your site is ready. It's funny how it [20:59] took some bots to help us improve a way [21:02] that we serve content to humans. [21:04] >> Yeah, exactly. [21:05] >> Ultimately, we're getting there. We're [21:07] getting there. Um, I really appreciate [21:08] all the actionable insights for how to [21:11] approach um what to refresh on the site [21:14] and and how that will actually impact [21:16] the user and the bots. I want to shift [21:18] gears to reporting and metrics a little [21:21] bit because all of this is great uh or [21:24] great ideas. However, we we have to take [21:26] our organizations with us and get that [21:27] buy in. So, Darcy, I I really want to [21:29] throw it to you first around um you [21:32] know, how how have you been framing the [21:35] ROI of investing in improving websites [21:39] and improving our marketing for this AI [21:41] world? You know, when a CFO asks [21:43] something like, "What is the the ROI of [21:45] this investment here?" Um how do you [21:48] talk about it? What metrics do you pull [21:50] in? Um that type of thing. Yeah, this [21:53] this this is the biggest um thing [21:56] hitting marketers right now, which is um [22:00] I I joke about this all the time. I I I [22:02] show my age, but I was in marketing [22:04] before digital and um then digital came [22:07] into the world and it was like a drug. [22:10] It's it is we are all addicted to this [22:12] drug called attribution because for the [22:15] last 30 plus years, we have been able to [22:17] directly attribute. Someone clicked on a [22:19] thing, they came to our site and they [22:20] bought and they bought at this level and [22:22] I can do all kinds of really beautiful [22:24] math to say that was worth it. This [22:26] place isn't worth it. Um, we are used to [22:29] it as marketers. Our CEOs are used to it [22:31] as leaders and our boards demand it. [22:34] Right? Show me that you are investing [22:37] your money in a smart way. [22:40] It's not possible anymore. The we all [22:43] have to wean ourselves off the drug. Um, [22:46] and the the reason this is going to be [22:47] so hard is we can't cold turkey it. We [22:50] can't just say we're not going to [22:51] attribute because still a lot of people [22:53] go through the traditional way. Like we [22:55] didn't just flip a switch and say [22:57] everything's going through chat GBT now. [22:59] There's still a large amount of of [23:01] activities that are coming through ads [23:03] and search and all the traditional [23:04] methods. So we have some that's [23:06] attributable, but the the portion that [23:08] is unattributable is going to grow and [23:10] grow and grow. And so we have to figure [23:12] out now how to paint a picture of what [23:15] is happening that shows that our money [23:18] and the way that we're spending it is is [23:20] working. So we've got our attributions [23:22] for LLMs on things like how many times [23:25] are we showing up, how many times are we [23:27] mentioned, how many times are we cited, [23:29] what's the sentiment of that that's [23:31] coming up. That's one way to say yeah, [23:33] that's sort of working. Then we've got [23:35] our traditional methodology around, you [23:38] know, clicks and that sort of thing. And [23:40] then we have our sort of traditional [23:42] brand and PR metrics. And you've got to [23:44] create this sort of mosaic that shows, [23:47] okay, this is how this is all starting [23:48] to come together. And um that's how we [23:51] figured out that like people coming in [23:52] from LLMs were converting faster. It's [23:54] not cuz we had direct traffic from LLMs [23:56] that we could track all the way through. [23:58] We just started painting a picture and [23:59] we're like, "Oh, interesting. We see [24:02] that we're starting to show up a lot [24:03] more here." And we're also starting to [24:05] see conversion rates um from direct [24:07] traffic suddenly spike. That's a [24:10] correlation. It's correlation not [24:11] directly attributable. But we have to [24:14] start to figure out what are those um [24:17] those indications that we can find in [24:18] our business that say the work that [24:20] we're doing is is is starting to pay [24:23] off. [24:24] >> I'm I'm also I also have the same [24:25] background. I I worked in an advertising [24:27] agency and I remember the advertising [24:29] agency uh advertising for its first uh [24:32] digital person and I think that was [24:35] literally the title at the time. So, so, [24:39] so yeah, I I I I completely agree with [24:41] what you're with with with what you're [24:42] saying in the sense that we are now [24:45] completely addicted to this click [24:47] brought me that. Uh, but yeah, that that [24:51] that ship is sailing. And to be honest, [24:54] now as a marketer, I completely regret [24:57] having those conversations with with [24:59] with the with with the boards and with [25:02] the people we reported to in the past [25:03] because now those people are used to it, [25:06] rightly so. And persuading them that [25:09] that chip has sailed is going to be a [25:10] really tough one because that made [25:12] sense. I I got 100 clicks and I got a 20 [25:14] conversions and I have proof that this [25:17] is working. It's hard to tell them, oh, [25:19] by the way, that that's not working [25:21] anymore. Uh but my favorite analogy [25:23] remains the billboards one. Uh I used to [25:26] work in in in in in a brand which [25:29] invested in a lot of billboards and my [25:31] most regular fight at the boardroom [25:33] level was but how do we know that the [25:36] billboards are bringing us sales? And my [25:39] response was always the same. Uh please [25:42] point me in the direction of the first [25:43] billboard you see where you see somebody [25:45] slotting money inside it. It's it's it's [25:48] not it's you can't attribute it what a [25:51] billboard for certain brands for digital [25:53] brands for example but for certain [25:55] brands billboards are really important [25:57] to keep your brand top of mind. So the [26:00] same applies in digital. So as as Darcy [26:03] was saying where where before we could [26:06] see something go from click to [26:08] conversion that is broken. I would add [26:12] something to that. Also keep in mind [26:15] that ad networks it is in their interest [26:19] to show you their ROI very strongly. So [26:24] it is in their interest to show you data [26:26] data which looks makes them look [26:28] amazing. So right now one big mistake [26:31] I'm seeing which is making all the CPCs [26:33] rise because everyone is falling into [26:35] this mistake is that everyone is buying [26:37] into oh ads give me a clear ROI so I'll [26:40] continue I'll I'll invest much more in [26:42] ads than I'll do in anything else that [26:44] is a mistake. It's not don't do ads, do [26:46] ads. But just because it gives you the [26:49] data, it feeds you the drug to use [26:51] Darcy's um analogy. Um it doesn't mean [26:56] it's now the only channel you should [26:58] look at. [26:59] >> Yeah. And I would say James, you know, [27:01] uh I love what you said is I wish we [27:03] wouldn't have gotten so into teaching [27:07] everyone about these direct metrics [27:08] because I think as a as an industry, as [27:11] a function, we have lost the art of [27:13] brand and how brand can can build [27:16] because you couldn't directly attribute [27:18] things like the billboards or or uh [27:21] something that you know the radio spot [27:23] or the podcast spot. It wasn't as [27:25] directly as attributable. And so we're [27:28] going to have to go back to the Madmin [27:29] days where we we recognize that those [27:33] things are important even if you can't [27:36] actually say in important to this you [27:39] know exactly this dollar amount worth um [27:42] and I think we're going to see it you [27:43] know if if you don't invest in your [27:45] content if you don't invest in the [27:47] things that are going to help you with [27:48] visibility um you will see your site [27:51] traffic is going to go down and your [27:53] business is going to go down right and [27:55] so you you are going to have to show [27:57] that you are showing up in these places. [27:59] It's just going to be a different set of [28:00] metrics that that start to show how much [28:02] value it brings. [28:04] >> If you could with in one sentence or a [28:06] couple of sentences give one more piece [28:08] of advice for every marketer who's [28:10] watching today when they go to their [28:12] next team meeting after this of what's [28:14] that one thing they should bring to the [28:16] team meeting. Um I would love for both [28:18] of you to take the opportunity to give [28:19] that. Uh Darcy, we'll start with you. If [28:21] if AI had to explain what we do, why we [28:25] matter, what would it say? And is it the [28:28] right thing? Uh pressure test that [28:30] answer. You know, um it it is the single [28:33] most we we have gotten so used to [28:35] content that's like about learning and [28:38] and structuring our website so that it's [28:40] about taking people through a path um [28:43] that we've probably gotten away from [28:45] that just so fundamental question which [28:47] is what do we do? Why does it matter? [28:49] How are we different? How are we [28:51] different from everyone else? Um because [28:53] AI is not going to know if if the words [28:56] you're saying on your page are the same [28:57] as as your competitor. It's not going to [29:00] know what to do if you haven't haven't [29:02] been really explicit. So I think it's it [29:04] really is that go pressure test who you [29:07] are, what's your value prop, what's your [29:08] differentiation, [29:10] um and see what's coming up. And I think [29:12] that'll be a good guide as to what you [29:14] might need to go um update pretty [29:16] quickly. [29:16] >> James, what about you? If you are a [29:18] marketer, you really need to take, if [29:21] you don't already, you really need to [29:22] take your customer experience and your [29:24] user experience much much more [29:25] seriously. I am a book nerd, so it would [29:28] be uh natural of me not to recommend the [29:31] book. This is uh an old one, a really [29:35] old one, and in fact, it it could do [29:37] with an update, but the principles are [29:39] still extremely solid, and I love its [29:42] title, and I recommend it. Anyone who [29:43] knows me for five minutes knows I [29:45] recommend this book and it's called [29:46] Don't Make Me Think. Uh I love the name. [29:50] Have a look at it. It's a wonderful [29:51] book. It honestly if that's the one next [29:54] action you do, that's already a huge win [29:56] in my view. [29:58] >> I love that. [29:59] >> Yeah. Well, thank you both so much. I've [30:01] learned a lot. I hope everyone else has [30:03] learned a lot. This has been uh really [30:05] really helpful. And thank you all again [30:06] for joining us today. Have a good one.  📄 **Full text:** ![[Full Document Contents/Articles/The Marketer’s Playbook for the Intelligent Web.md|The Marketer’s Playbook for the Intelligent Web by YouTube]] ## Highlights - in our inbox with newsletter after newsletter of like, we we know how to do it and you're behind and you should do it. Like the bottom line is like no one's got this figured out yet. This is still new. like with SEO, like did we ever fully figure it out? To your point, James, right? Like there's always something that's changing. So, for me, it's really like uh how do we how do we break this thing that's out there in the ether that we're all behind and late and we got to catch up? Like, we're all right where we're supposed to be right now. We're all learning. We're all ([View Highlight](https://read.readwise.io/read/01kr1w0bagr8qhwmv185xj4vws)) - Tags: 👻 ai highlighted - I'm sure is no strange question to you um going from company to company when you look at a website when do you know it's time to rebuild it um when it comes to this AI world of search I is rebuilding an important thing to ask yourself now and if so what do you usually expect the impact to be? Yeah, you know, a year ago, I I actually had this very conversation with some of my CMO colleagues because, you know, being in the website business, they're like, "What are you seeing? What do you think? What do we need to do?" And um a lot of ([View Highlight](https://read.readwise.io/read/01kr1w0bba8kcwy68mwh2dmea2)) - Tags: 👻 ai highlighted - you to know the things that I started looking for on our site are things like you know is your traffic flat or down but the industry demand isn't right this is that zero you're getting that zero click phenomenon you know you're creating a lot of content but it's not showing up or it's not getting cited. Uh is your site or still organized around pages versus topics and expertise. Um you know it it's just like you know can you move fast? Are you able to make changes really really quickly? So that ([View Highlight](https://read.readwise.io/read/01kr1w0bcdnvpg9m6vvgdf2n69)) - Tags: 👻 ai highlighted - that we serve content to humans. >> Yeah, exactly. >> Ultimately, we're getting there. We're getting there. Um, I really appreciate all the actionable insights for how to approach um what to refresh on the site and and how that will actually impact the user and the bots. I want to shift gears to reporting and metrics a little bit because all of this is great uh or great ideas. However, we we have to take our organizations with us and get that buy in. So, Darcy, I I really want to throw it to you first around um you ([View Highlight](https://read.readwise.io/read/01kr1w0bgytkvpszjbcpy6j62q)) - Tags: 👻 ai highlighted - know, how how have you been framing the ROI of investing in improving websites and improving our marketing for this AI world? You know, when a CFO asks something like, "What is the the ROI of this investment here?" Um how do you talk about it? What metrics do you pull in? Um that type of thing. Yeah, this this this is the biggest um thing hitting marketers right now, which is um ([View Highlight](https://read.readwise.io/read/01kr1w0bjxv244sdz024k8f8z3)) - Tags: 👻 ai highlighted - answer. You know, um it it is the single most we we have gotten so used to content that's like about learning and and structuring our website so that it's about taking people through a path um that we've probably gotten away from that just so fundamental question which is what do we do? Why does it matter? How are we different? How are we different from everyone else? Um because AI is not going to know if if the words you're saying on your page are the same as as your competitor. It's not going to ([View Highlight](https://read.readwise.io/read/01kr1w0bscax4njvx71febysgw)) - Tags: 👻 ai highlighted