Making Drupal Work - AI Search and Personalization Webinar

This webinar explains how Drupal sites can improve AI search visibility and personalization by structuring content, adding schema, improving metadata, activating visitor data, and measuring AI citations, engagement, and task completion. It gives teams a practical 90-day roadmap to make Drupal content easier for AI and users to find.

Transcript

1. Making Drupal Work in the Age of AI Search and Personalization Date : May 19th, 2026 | 11:00 AM PST / 2:00 PM EST | Live on Zoom Karthik Kalimuthu | Chief Technology Officer (CTO) John Harris | Vice President (VP)
2. Your Speakers Today Karthik Kalimuthu Chief Technology Officer, BinaryWorks • 18+ years architecting government and enterprise Drupal environments • Led WCAG audits across city, county, university, and public library Drupal sites • Specializes in accessibility remediation and Section 508 compliance • 50+ public sector organizations on ADA Title II compliance strategy Shawn Post Head of Sales, BinaryWorks • 13+ years supporting Drupal solutions across government and enterprise environments • Leads sales strategy for Drupal accessibility and compliance initiatives • Specializes in ADA, WCAG, and Section 508 compliance solution planning • Works with public sector organizations to align technology solutions with compliance and accessibility goals
3. It's Already Happening A prospective student asks ChatGPT: "Which US universities offer a part-time public health master's with an online option?" Your program exists. Your page has the answer. Two competitors are cited. You're not. That student builds a shortlist of three. You're not on it. Similarly, • When a citizen searches for Section 8 eligibility • A donor researches nonprofits addressing food insecurity - the same thing happens. Your content exists. Not a content problem. A structure problem. And it's fixable.
4. Two Gaps. One Platform. Before They Arrive The Gap AI engines skip your content - a competitor or third- party gets cited instead The Cost You're absent from the decision before a visit ever happens When They Arrive The Gap Every visitor sees the same experience regardless of who they are. The Cost Buried content. Abandoned tasks. Rising support burden. Both trace to the same root: a site built around how your organization is structured - not how your users find and use information.
5. # What We Cover 1 The problem your users are having 2 Diagnose your Drupal site 3 Fix 1 - Make your content findable 4 Fix 2 - Serve each visitor differently 5 Implementation toolkit 6 Measure and prove it's working 7 90-day plan What This Session Covers
6. 01 The Problem The Problem Your Users Are Having
7. Where Your Users Go Before They Reach You High-intent research queries - the ones that shape shortlists - go to AI first. What Your User Is Asking Where They Ask Who Gets Cited "Best US universities for environmental policy research" ChatGPT, Perplexity Universities with structured, question- answering program pages "How to apply for Section 8 housing in [state]" Google AI Overview Third-party guides - not the agency page with the application "Most effective US nonprofits working on urban food insecurity" Perplexity, ChatGPT Organizations with entity-clear, structured web content "Grants available for rural community health clinics" ChatGPT, AI Overview Whoever has the most structured, extractable answer "How to appeal a Medicaid denial in [state]" Google AI Overview Advocacy orgs with structured FAQ pages - not the state portal The pattern: your content exists. AI can't parse it clearly enough to cite it. The user moves on.
8. Find Out Where You Stand Right Now 1 Open ChatGPT or Perplexity 2 Type the query your ideal prospective student, citizen, donor or your user would ask to find you 3 Are you in the answer? Who is instead? 4 Look at what cited content has that yours doesn't - structure, specificity, directness If you're not cited - you didn't lose at ranking. You were absent before the decision even started. That gap exists for every query your audience asks about your programs, services, and cause areas.
9. One Site. Every Visitor Gets the Same Page. Who Arrives What They Need What Your Site Shows Prospective international grad student Program details, visa info, scholarships, how to apply Same homepage as a current domestic undergrad Parent researching for their child Cost clarity, financial aid, student support services Admissions page written for the applicant Citizen with a specific benefit need One answer, one next step Dept directory organized by agency name - not by task Long-term donor Impact evidence, updated programs, giving options Same homepage as a first-time visitor Returning volunteer Upcoming opportunities, their history Identical experience as someone discovering the org for the first time Every visitor who doesn't find what they need in the first two interactions is a support call, a lost application, or a donor who gave elsewhere.
10. What Both Gaps Are Costing The Gap The Real Business Cost Not cited in AI search Prospect, donor, and citizen shortlists built without you on them Generic experience on arrival Every incomplete task becomes a support call your staff handles manually Content published but not found Publishing investment producing content your audience cannot discover No AI visibility tracking Ground is lost before it shows in data - as drops in applications or retention Ask yourself: How many support tickets, calls, or emails last month answered questions already on your site? That number is the cost of both gaps - made visible.
11. The Window Is Open - But Not Permanently AI citation is earned. It's not paid for. 165x Gen AI traffic is growing 165x faster than organic search traffic WebFX, 2025 • Organizations that structure content now will own citation authority - durable and hard to displace • Your Drupal platform is natively capable of everything in this session • You're not rebuilding - you're activating what the platform already makes possible
12. Score Your Site Right Now Score each signal: 1 = Not in place 2 = Partial 3 = Fully in place Dimension Signal Score Findability: Structure Key content in typed fields - not body blocks /3 Findability: Structure IA organized around user tasks - not your org chart /3 Findability: Signals Every page has unique metadata from field-level tokens /3 Findability: Signals Schema markup implemented and validated /3 Experience: Personalization Different experiences served to distinct visitor roles /3 Experience: Personalization Segments defined from behavior and role /3 Experience: Measurement Engagement depth tracked - not just page views /3 Experience: Measurement AI citation visibility monitored separately from traffic /3 Maximum Score: 24
13. 500+ Drupal AI Readiness Audits. Same Problems. Every Time. What We Find How Often Zero schema markup of any kind 9 in 10 sites Metadata auto-generated - not field-configured 8 in 10 sites Critical content in body fields - no structure 8 in 10 sites Identical experience served to every visitor 9 in 10 sites First-party visitor data collected, never activated 7 in 10 sites No way to track AI search citation 6 in 10 sites If your scorecard reflects this - you're not the exception. You're the pattern. Book your free Drupal AI Readiness Audit - Find out which gaps your site has
14. What Your Score Tells You Score What It Means Where to Start 6-10 Significant gaps across both dimensions Content architecture + Tier 1 schema - Week 1 11-16 Foundations in place. Specific gaps are actively costing you. Target lowest-scoring dimension first 17-24 Ahead of most Drupal implementations Close remaining gaps. Sharpen measurement. Every dimension you scored below 3 has a specific fix in the next two chapters.
15. 02 Fix 1 Fix 1: Make Your Content Findable Your content exists. AI can't parse it, extract it, or cite it.
16. How AI Decides What to Retrieve and Cite AI doesn't rank pages. It runs a three-step pipeline - and most Drupal sites drop out at Step 1. Step What AI Does Where Drupal Sites Drop Out Retrieve Pulls sources it can parse based on query intent Body-field content is one opaque block - deprioritized Synthesize Extracts a clean answer from labeled, structured content No labeled sections means no clean extraction Cite Attributes sources that are specific, schema- marked, and authoritative No schema + generic metadata = skipped Try this: Open a key page on your site. Strip the navigation and the page title. Can AI tell what this page is, who it's for, and what specific question it answers? On most Drupal pages - it can't. That's the retrieval gap.
17. Fix 1A - The Body Field Problem Most Drupal teams default to body fields. To AI, a body field is one unstructured block - no labels, no hierarchy, nothing it can reliably extract and cite. Body Field Structured Fields Program details in long prose Discrete fields: delivery format, credit hours, deadlines, tuition Service eligibility buried in paragraphs Labeled fields: who qualifies, documents required, how to apply Nonprofit impact in embedded PDFs Web-native fields: outcomes, beneficiaries served, geography Policy as unstructured prose Fields: applies-to, effective date, key provisions, contact Structured fields produce automatic schema output and give personalization rules clean signals - one architecture decision, two outcomes.
18. Fix 1B - Schema Markup: Tell AI What Your Content Is Without schema, AI guesses what your content means. With schema, you tell AI explicitly what the page is, who it's for, and what it contains. 2.5x More likely to appear in AI-generated answers 40% More AI Overview appearances with Tier 1 schema complete Schema helps LLMs interpret and cite content accurately (Microsoft, SMX Munich 2025) Implementation in Drupal - no custom development required for Tier 1: • JSON-LD output via Metatag module + Token system • Schema properties map automatically from content type fields • Configure once per content type - scales across every page of that type Validate every sprint: Google Rich Results Test
19. Schema Priority by Sector Tier Schema Type Apply To Tier 1 - This Month Organization Homepage, About FAQ Page Service, program, and resource pages Article + Person All authored content with expertise signals BreadcrumbList Site-wide Tier 2 - Next Quarter EducationalOrganization + Course All degree and certificate program pages GovernmentOrganization + GovernmentService All public-facing service and benefit pages Nonprofit + Event + SpecialAnnouncement Program, event, and funding pages Implementation: Metatag module + Token system across all schema types.
20. Fix 1C - Metadata That Actually Works What most Drupal sites have: auto-generated metadata from page title and first 160 characters of body content. Identical across pages. No entity signal. Nothing citable. Element What's Broken What Works Meta title Node title - often a dept or system label [Program Name] - [Format] - [Institution] Meta description Auto-excerpt from body Summary field: who it's for, what it covers, next step URL /node/4827 /services/housing/apply-section-8 Open Graph Site default or missing Per content type: relevant image + descriptive field Fix: One Metatag template per content type. Token-based output from discrete fields. Scales across thousands of pages without manual work per page.
21. Fix 1D - Write Content That Gets Cited The shift: from publishing about topics to owning the answer to a specific question. What Most Sites Publish What AI Cites "Graduate Admissions" overview page "How to Apply for Graduate Admission - Requirements, Deadlines, and Steps" "Citizen Services" department directory "How to Apply for Emergency Rental Assistance - Eligibility and Documents" "Our Programs" nonprofit overview "What Grants Does [Org] Offer for Rural Health Clinics" Prose service descriptions FAQ blocks with direct answers to real user questions The GEO Formula for every key page: 1. First 100 words: what this page is, who it's for, what it answers - explicitly 2. H2/H3 headings: written as your user asks - not as your team names things 3. FAQ block: 3-5 questions from on-site search data and support ticket logs 4. Keep content current: AI weights recency - outdated pages lose to fresher third-party sources
22. Case Study: A Renowned University Challenge • Program pages built on body fields - zero schema, auto-generated metadata across 80+ pages • Absent from AI answers while competitors with weaker programs were being cited What Changed • Restructured content types with discrete fields; EducationalOrganization + Course schema via Metatag • Field-level Metatag templates per content type; FAQ paragraph type on 80+ program pages • Questions sourced from on-site search data and admissions support ticket logs Results +67% AI citation rate on target program queries +34% Engagement rate on program pages +28% Admissions inquiry form completions
23. Quick Fixes You can Implement: Findability This Week • Search top 5 user queries in ChatGPT + Perplexity. Are you cited? Document your baseline. • Test key pages at search.google.com/test/rich-results. See what's there and what's broken. This Month • Implement Tier 1 schema on 5 highest-traffic pages via Metatag + Token • Configure field-level Metatag templates per content type - replace auto-generated output • Add FAQ blocks to top 10 pages using on-site search data and top support ticket topics Next Quarter • Audit body-field pages. Migrate structured content to discrete, labeled fields. • Expand to Tier 2 schema for your sector. Validate with Rich Results Test after each sprint.
24. 03 Fix 2 Fix 2: Serve Each Visitor the Right Experience Fix 1 gets them to your site. Fix 2 determines whether they find what they came for, complete their task, and return.
25. Why Every Visitor Gets the Same Page This isn't a design oversight. It's the natural outcome of three common decisions: Content organized by department not by what visitors are trying to do No segment definition the site doesn't distinguish between a returning donor and a first-time visitor First-party data collected but never activated visitor signals sit unused in every Drupal implementation we audit None of this requires a rebuild. It requires deliberate configuration - starting with what you already have.
26. The Data Already Sitting in Your Drupal Site Before adding any new tool - audit what you already have. Signal What It Tells You Where to Find It On-site search queries What users couldn't find through navigation - in their own words Search API / site search logs Content category affinity Which topics each visitor engages with across a session GA4 + content grouping Return visit patterns Who comes back, how often, and what they do differently GA4 user behavior reports Authenticated vs. anonymous behavior How staff, students, or members differ from first-time visitors Drupal roles + session data Form interaction history Where key journeys start but don't complete GA4 conversion events The data for your first three personalization use cases is almost certainly already here. The gap is activation - not collection.
27. Three Segments to Start - In Order Segment Logic First Action Role-based Authenticated user vs. anonymous visitor Different homepage block: task shortcuts for returning users, orientation content for first-timers Behavior-based Deep session (4+ pages, clear topic affinity) vs. surface visit Surface a specific next step to high-intent visitors who haven't taken action yet Journey-based Started a key task and didn't complete Re-engagement block on return visit targeting the incomplete action Build role-based first. Lowest effort. Clearest data signal. Prove value - then expand.
28. Where to Personalize First Not every page. Start where audience divergence is widest. Touchpoint Why High-Impact What to Personalize Homepage Every segment lands here Hero content, featured resources, CTA by role Service or program discovery Routing friction is costliest here Pre-surfaced content based on role or prior behavior Resource library Manual navigation kills engagement depth Topic-based surfacing from content category affinity Authenticated portal entry Role-based expectation is highest Dashboard content, announcements, role-specific next actions High drop-off pages Greatest recovery opportunity Simplified next step, contextual support content
29. Questions Your Site Should Be Able to Answer If any of these is "no" - you have an immediate use case. 1 Does a donor who gave three times last year see anything different from someone who just discovered your organization? 2 Does a student in their final semester see a different homepage than a first-week freshman? 3 Does a citizen who already submitted an application get a different experience than one who hasn't started? 4 Does a returning visitor who spent time on your grant programs page get a relevant next step on their second visit? 5 Do you know what content your most engaged visitors read before they took action - and are you using that pattern for anyone else?
30. Case Study: A Nonprofit Organization Challenge • Single platform serving donors, applicants, beneficiaries, and volunteers - identical experience for all • High support contact volume for questions already on the site What Changed • Role-based personalization via Smart Content module - authenticated vs. anonymous homepage differentiation • Behavior-based content surfacing for visitors with 4+ pages of program engagement • FAQ blocks on top 15 service pages using support ticket backlog as the question source Results -41% Support tickets for site-answerable questions +38% Engagement depth - authenticated segment +26% Program application completion rate
31. Quick Fixes: Personalization This Week • Homepage: does anonymous differ from authenticated? If not - set up one block visibility rule. Measure for two weeks. • Pull 90-day on-site search data. Top missed queries = your personalization brief. This Month • Configure GA4 content grouping by topic category - foundation for behavior-based segmentation • Install Smart Content. Build role-based segment. Serve different primary CTA per role on the homepage. • A/B test one personalized experience vs. default on your highest-traffic page. Run 30 days. Measure task completion. Next Quarter • Build behavior-based next-best-content rules for visitors with 3+ pages of topic engagement • Activate journey-based re-engagement blocks for users who started but didn't complete a key task
32. Implementation Toolkit: Findability Category Tool Use Drupal Modules Metatag + Token Schema output and field-level metadata configuration Pathauto Semantic, intent-based URL generation Search API Structured indexing and on-site search Schema.org Metatag Extended schema type support Validation Google Rich Results Test Schema validation per page Google Search Console AI Overview impression tracking AI Monitoring Otterly.AI / Profound Automated citation tracking across AI engines Content Research AnswerThePublic Mine real user questions for FAQs and headings
33. Implementation Toolkit: Experience Category Tool Use Drupal Modules Smart Content Condition-based display - role, behavior, location Block Visibility Groups Fast rule-based block display by context Acquia Personalization Advanced segmentation + A/B testing Rules module Complex multi-condition content routing Analytics GA4 + Content Grouping Foundation for behavior-based segmentation Microsoft Clarity Scroll depth and task drop-off identification Testing Acquia Optimize A/B test management on personalized experiences Tier 1 fixes - Metatag, Smart Content, Block Visibility - require configuration, not development. Start there. Add complexity only when the first use case proves value.
34. One Decision. Both Problems Solved. The architecture decision that makes your site findable by AI also makes it personalizable. Both require content organized around user need - not internal structure. The Decision Impact on Findability Impact on Experience Content in discrete typed fields Schema is automatic. AI extracts cleanly. Personalization rules have specific, reliable signals. IA built around user tasks Question-based content earns citation. Right content surfaces to the right segment naturally. Taxonomy-driven organization AI retrieval is consistent across topics. Segment routing requires no complex logic. First-party data activated Reveals which content to prioritize for AI. Powers behavior and journey-based segmentation. You're not running two workstreams. One investment. Two outcomes. This is Drupal's structural advantage - when it's built right.
35. 04 Measuring Measuring What Matters
36. The Metrics Hiding the Problem What Most Teams Track What It Misses Total traffic volume AI citations that shaped a decision before - or instead of - a visit Page views Whether visitors completed the task the page was built for Bounce rate Whether visitors who stayed engaged or just loaded the page Organic rankings Whether your content is cited in AI answers regardless of rank Overall session count Whether personalization produces different outcomes per segment The result: dashboards reporting stability while consideration share and task completion quietly decline. Problems compound and surface months later - as a drop in applications, a support spike, or declining donor retention.
37. Tracking AI Visibility Frequency Action Tool Weekly Search top 5 user queries in ChatGPT + Perplexity. Cited? Changed? Screenshot and document. Manual - 15 min Monthly Review AI Overview impressions for schema-implemented pages Google Search Console Monthly Monitor branded search volume - AI citation produces branded search lift in 60-90 days GA4 Ongoing Automated citation tracking across ChatGPT, Perplexity, and Google AI Overviews Otterly.AI / Profound Run the manual check before any schema implementation. The before/after delta is your proof of impact.
38. Tracking Engagement Depth Replace traffic volume with signals that show whether the site actually worked for the visitor. Metric What It Measures Tool Task completion rate Whether visitors complete the action the page is built for GA4 conversion events Engagement rate Sessions with meaningful interaction - not just pageloads GA4 Support ticket volume Whether the experience gap is closing - most honest signal available CRM / help desk Return visit frequency Whether engaged visitors come back GA4 user reports Personalization-specific: • Segment-level engagement vs. default - does the personalized segment outperform? • A/B test results: personalized vs. default on the same page • Track before/after on specific journeys when personalization rules are introduced
39. 05 Your Plan Your Path Forward
40. Phase 1 - Audit and Foundation Weeks 1-2 | Establish baselines. Start the highest-leverage fixes. Findability • Run the scorecard against your live site in detail • Search top 5 user queries in ChatGPT + Perplexity - document citation baseline • Test key pages with Google Rich Results Test • Implement Tier 1 schema on top 5 highest- traffic pages Experience • Pull 90-day on-site search data. Identify top missed queries. • Identify first personalization use case. Define three starting segments. Deliverable: Citation baseline documented. Tier 1 schema live on 5 pages. First use case defined.
41. Phase 2 + Phase 3 - Activate and Optimize Phase 2 - Activate | Weeks 3-6 Findability Add FAQ blocks to top 10 pages. Configure field-level Metatag per content type. Experience Deploy use case one: role-based homepage personalization. Set up GA4 engagement events. Deliverable: 10 pages with FAQ + schema. Metadata configured. Personalization use case one live. Phase 3 - Optimize | Weeks 7-10 Findability Compare AI citation rate vs. baseline. Restructure top 10 body-field pages. Expand Tier 2 schema. Experience A/B test personalized vs. default on highest-traffic page. Define use case two from results. Deliverable: Tier 2 schema live. A/B results in hand. Body-field pages restructured. Use case two defined.
42. Phase 4 - Scale and Embed Week 11 onwards | Make this operational - not a one-time project Cadence Action Weekly Search 5 priority queries in ChatGPT + Perplexity. Screenshot. Document changes. Monthly Review Search Console AI Overview data. Update schema where gaps appear. Monthly Review support ticket volume for site-answerable questions. Track the trend. Quarterly Re-run the readiness scorecard from today. The delta is your proof of progress. Ongoing Freshness protocol: content update triggers when eligibility, deadlines, or program details change
43. Five Things That Are True After This Session 1 You're being bypassed before anyone visits your site. AI search is where your audience builds their shortlist. Not cited = not considered. 2 The body field is your most common and most fixable structural problem. Moving content to discrete fields is the highest-leverage change most teams can make. 3 Personalization is an activation problem - not a budget problem. The signals and modules needed for your first three use cases exist in your Drupal environment right now. 4 Findability and personalization share one foundation. Content architecture decisions that improve AI citation also enable segmentation. One investment. Two outcomes. 5 Your support ticket volume is your most honest performance signal. When it drops for site-answerable questions - the platform is working.
44. Find Out Where Your Drupal Site Stands What We Audit What You Get Content structure analysis Field vs. body - page by page on your live site Schema + metadata audit Gap report with Tier 1 priorities, sector-specific Personalization readiness Signals available, segment possibilities, module fit Measurement maturity Current tracking vs. what's needed Prioritized recommendations Ranked by business impact and effort - specific to your site One session. Full audit. Clear priorities. No obligation. Book Your Free Drupal AI Readiness Audit
45. Thank You Q&A BinaryWorks(Formerly DrupalPartners) | www.binaryworks.com Drop them in the chat now