Invisible to AI search. Six operational gaps nobody had mapped. A 6-week content project became a 9-month engagement.
A $14B wealth advisory firm hired Work-Smart to fix AI search invisibility. The engagement uncovered six operational problems that had nothing to do with the website. What started as a content project became a 9-month Fractional Head of AI engagement, proposal automation, knowledge systems, intelligence briefing, and team training. Everything built belongs to the firm.
| Client | $14B wealth advisory firm, ultra-high-net-worth families, South Florida |
| Starting problem | Invisible to AI search engines despite 12 years of published expertise |
| What it became | 9-month Fractional Head of AI engagement across 6 operational areas |
| Investment | Monthly retainer. 90-day trial. 9-month engagement |
| Key deliverables | Voice DNA Profile, 60 authority pages, proposal engine, knowledge base, intelligence briefing, team AI environments, Operational Blueprint |
| Outcome | Every system built, documented, and owned by the firm. No ongoing dependency |
The Starting Problem. AI Invisibility
A $14B wealth advisory firm, one of the most respected independent wealth advisors in the United States. Fee-only, non-discretionary, fiduciary. Serving ultra-high-net-worth families. More than a decade of published expertise across investment strategy, governance architecture, family enterprise frameworks, and market commentary. A deep podcast library featuring their partners.
When a family searched the firm by name, it appeared. When they asked "how to evaluate a wealth advisor" or "what to look for in a fee-only fiduciary", questions the firm had answered comprehensively across dozens of documents, the firm was nowhere.
The data: roughly two orders of magnitude between branded and unbranded search performance. Strong branded CTR, near-zero non-branded CTR. A deep library of published documents that AI search engines couldn't read. Podcast episodes without transcripts, invisible to any model that can't listen to audio. Meta descriptions too long. Pages missing H1 tags. A sitemap from 2013, triplicated across Google and Bing webmaster tools.
The firm's marketing agency had been doing solid work on the website itself. But none of it was structured for the shift already happening: the migration of search behavior from Google to AI engines like ChatGPT, Claude, and Perplexity.
Competitors were already restructuring their content strategies around AI-readable question-and-answer frameworks. The firm had better content. They just hadn't made it readable by the machines.
Phase I-III. AI Visibility
Voice DNA Extraction
The first deliverable was not content. It was a 220-line Voice DNA Profile, a structural blueprint of how the firm communicates, derived from analyzing every document published since 2013. The profile captures argument patterns, vocabulary precision, tone calibration, and the things the firm never does: no exclamation points, no first-person storytelling, no client testimonials, no performance data, no urgency language. This became the calibration layer for all AI-generated content. Every page passes a 10-point writing test before it ships.
Authority Framework
10 questions the firm needs to own in LLM responses, derived not from keyword tools, but from partner interviews, client onboarding processes, and the frequently asked questions document the firm maintains internally. Each question maps to an authority pillar. Each pillar generates multiple pages approaching the same core question from different angles, industry contexts, and family scenarios. AI models learn through structured repetition. One page doesn't register. Twenty pages, all internally linked, all in the firm's authentic voice, that's what makes the model cite the firm.
60 Authority Pages + Technical Foundation
60 authority pages delivered. Each page: minimum 1,300 words, structured for machine readability, written in the firm's calibrated voice, with proper schema markup, question-and-answer formatting, and internal linking to the hub page. Sitemaps consolidated. Missing meta descriptions and H1 tags fixed. LinkedIn profiles reviewed with specific recommendations for each partner. Podcast transcript production plan established for all 77 episodes.
The AI visibility project delivered what it was designed to deliver. But the conversations with each partner revealed something the website project couldn't solve.
The Six Problems
Every Proposal Starts From Zero
Partners spend 4-8 hours building each proposal from scratch. The firm has produced hundreds of proposals over 12 years, but none of that institutional knowledge feeds into the next one.
What should happen: An AI-powered proposal engine that generates an 80% first draft in minutes, pulling from the firm's history of accepted proposals, fee structures, and service descriptions.
Knowledge Lives in Unsearchable Places
12 years of institutional knowledge, investment memos, meeting notes, research reports, client communications, scattered across drives, inboxes, and individual machines. When a partner needs something, they ask the person who might remember.
What should happen: A structured SharePoint knowledge base where any document from 12 years is retrievable in seconds. Not a file dump, a searchable, tagged, AI-indexed institutional memory.
Email Contains 90% of Client Data
The relationship with 105 ultra-high-net-worth families lives primarily in Outlook. Client preferences, family dynamics, investment concerns, meeting follow-ups, all trapped in individual inboxes.
What should happen: A structured evaluation of CRM and client intelligence tools, starting with what the firm already owns (Microsoft ecosystem) before recommending external platforms.
Tools Deployed But Not Configured
The firm pays for Copilot, SharePoint, Power Automate, Copilot Studio, and Addepar. Most are partially deployed. None are configured to work together or optimized for how the firm actually operates.
What should happen: Every tool evaluated, configured for the firm's specific workflows, and connected where possible. Not new purchases, activation of what already exists.
Meeting Transcripts Unused
Teams records and transcribes every meeting. The transcripts sit in a folder. Nobody reads them after the meeting. Years of client conversations, partner discussions, and strategic decisions, unstructured and unsearchable.
What should happen: Structured intelligence extraction: every concern, commitment, priority, and follow-up pulled automatically from transcripts and organized by client, topic, and date.
Intelligence Briefing Is Manual
Partners monitor 15-20 sources every morning, market data, news, client-relevant events, regulatory changes. This takes 2+ hours and produces no shareable output.
What should happen: One briefing, one place, shareable across the team. Automated aggregation from all sources, filtered by relevance to the firm's clients and investment thesis.
Phase IV. Fractional Head of AI
The AI Visibility work proved two things: the methodology works, and the firm has operational problems that go far beyond the website. A Fractional Head of AI works inside the firm, not as a consultant delivering recommendations, but as an operator building systems. Everything built belongs to the firm. 30 days notice, no lock-in.
| Model | Fractional Head of AI |
| Duration | 9 months (3 phases of 3 months each) |
| Investment | Monthly retainer. 90-day trial. 9-month total engagement |
| Structure | Prove → Build → Embed |
| Exit | 30 days notice. No lock-in. Goal is self-sufficiency |
90-Day Trial
Configure the AI stack (Copilot, SharePoint, Power Automate). Build personal AI environments for each partner. Deploy Lyra analysis for client intelligence. Deliver the Operational Blueprint, a documented assessment of every operational gap, every tool, every workflow, with specific recommendations and build timelines. The firm decides whether to continue based on evidence, not promises.
Blueprint Becomes Working Systems
The Operational Blueprint becomes the build plan. Proposal engine. Knowledge base. Intelligence briefing. Meeting transcript extraction. Each system built, tested, and documented. Weekly briefings with partners. Hands-on training sessions with the team.
Transition to Self-Sufficiency
Everything transitions to the firm. Systems documented. Team trained. Workflows embedded. The goal is that when the engagement ends, the firm operates these systems independently. If they want a maintenance retainer, that option exists. But the dependency ends.
Before and After
| Area | Before | After |
|---|---|---|
| Proposals | 4-8 hours from scratch | 80% first draft in minutes |
| Institutional memory | Depends on who remembers | Any document from 12 years retrievable in seconds |
| Partner's morning | 2+ hours across 20 platforms | One briefing, one place, shareable |
| Meeting prep | Hours of manual assembly | Every concern, commitment, priority pulled automatically |
| Technology spend | Most tools not activated | Every tool evaluated and configured |
| AI readiness | No training, no governance | Weekly briefings, workshops, personal environments |
| Institutional continuity | Partner leaves, knowledge leaves | Firm's knowledge structured and retained |
| Future decisions | Based on vendor pitches | Grounded in Operational Blueprint built from evidence |
What This Costs in Context
| Option | Cost |
|---|---|
| This engagement | Monthly retainer, 90-day trial |
| AI consulting retainer | $20K-$40K/month |
| Full-time AI Director | $200K-$300K+/year |
| Lyra CRM | $100K-$300K+/year |
| Value of 1 new family | $500K+/year |
9
months
90-day
trial
6
operational gaps solved
Why This Story Matters
This engagement didn't start with a 9-month proposal. It started with a specific, contained problem: the firm was invisible to AI search engines. The AI Visibility work was scoped at 6 weeks. It delivered what it promised.
But the process of building AI visibility required understanding how the firm operates , how partners create content, where knowledge lives, how client relationships are managed, what tools exist and how they're used. Those conversations surfaced six operational problems that no one had mapped. Not because people weren't aware of them, but because no one had looked at the firm's operations through the lens of what AI can now do.
That's the pattern. Start with what's visible. Fix it. Let the deeper questions surface. Build the next phase based on evidence from the last one. The firm never committed to 9 months on day one. They committed to 6 weeks, saw the results, and decided what came next based on what they learned.
Questions About This Case Study
The AI Visibility work required deep conversations with each partner about how the firm creates content, manages knowledge, and communicates with clients. Those conversations revealed operational problems, proposal automation, knowledge retrieval, meeting intelligence, tool configuration, that had nothing to do with the website. Each phase uncovered the next.
A Fractional Head of AI who works inside the firm. Not a consulting team that delivers a deck. One person who reviews operations, configures AI tools, builds custom systems, trains the team hands-on, and delivers an Operational Blueprint. Everything built belongs to the firm. 30 days notice, no lock-in.
A competitor did exactly that, at full salary plus benefits. This engagement delivers the same output on a monthly retainer, with no permanent headcount, no recruiting risk, and a 90-day trial before full commitment.
The dependency ends. The systems stay. The knowledge stays. The team is trained. The Operational Blueprint is documented and executed. If the firm wants to continue with a maintenance retainer, that option exists. But the goal is self-sufficiency, not ongoing dependency.
Yes. The approach is the same: start with what's visible (usually AI search or data chaos), fix it, and let the deeper operational questions surface. The 90-day trial structure means you see proof before committing.
The same model has been applied to construction, legal, packaging manufacturing, and nonprofit consulting. The operational problems are different. The phase-based approach is the same.
If your firm has years of expertise that AI search engines can't find, or operational problems that no one has mapped, you're in the same position this firm was before the first conversation.
Every engagement starts with a contained first step. You see results before deciding what comes next.