Prospect
What prompted interest, and what should the first conversation cover?
Context brief, agenda, qualification questions, follow-up draft.
RIA Client Journey Intelligence Playbook
Map the client journey before you choose the AI tool.
A visual field guide for boutique RIAs that want AI help across prospecting, onboarding, discovery, planning, IPS work, review meetings, follow-up, and ongoing relationship management.
Founder-led RIAs
Complex advisory operations
Fiduciary-first teams
The problem
Tool-first AI creates client-trust risk and missed-opportunity risk at the same time.
Tool-first rollout
Staff experiment in different tools, vendors pitch copilots, and leadership cannot clearly answer what client data is being used, who reviews outputs, or where the decision record lives.
Governed adoption
The firm connects AI to client-service workflows, protects client information, keeps advisor judgment visible, and gives counsel and IT clear rules to work from.
Client journey intelligence
The client journey becomes the map for AI decisions.
The firm already has the client context. It is just split across CRM, planning, portfolio analytics, custodians, risk tools, IPS documents, Microsoft 365, SharePoint, email, and meeting notes. Joans helps organize the work around the journey instead of adding another place for advisors to check.
Onboarding
What data, accounts, authorizations, and expectations need to move cleanly?
Checklist, missing-item tracker, handoff summary, client-message draft.
Discovery
What do we know about goals, constraints, values, family context, and risk?
Discovery synthesis, open questions, values/risk/profile summary.
Planning
Which assumptions matter, and what still needs advisor judgment?
Assumption checklist, change summary, review packet, gaps to resolve.
IPS
Does policy language reflect objectives, restrictions, values, and risk?
Draft support, redline summary, drift questions, approval tracker.
Implementation
Does the proposal line up with plan, IPS, values, restrictions, and reality?
Proposal context, alignment checks, exception list, advisor notes.
Review meetings
What changed, what needs attention, and what should be raised with the client?
Review packet, agenda, open follow-ups, drift review, client-service memory.
Follow-up
What promises were made, who owns them, and what needs to be finished?
Follow-up draft, task support, handoff summary, overdue-item review.
Life events
What changed in the client life context, and what workflows are affected?
Event summary, affected-workflow checklist, advisor questions, sensitivity flags.
Relationship management
Is the relationship current, responsive, and ready for the next interaction?
Relationship memory, stale-item review, next-prep prompts, cadence support.
Three-lens model
Start with the work, not the software.
A useful AI roadmap looks at the firm from three angles at once: what people do every day, how the business runs, and how clients experience the firm.
Day-in-the-life
Where advisors, client-service teams, operations leads, and partners lose time every week: meeting prep, follow-up, inbox work, research intake, notes, and recurring requests.
Business operations
Where the firm needs repeatable handoffs: approvals, vendor work, reporting, compliance review, billing support, onboarding, and team knowledge.
Client journey
Where client experience improves or breaks: first response, meeting quality, planning follow-through, reporting clarity, and the moments when advisor judgment must stay visible.
Advisor workflow intelligence
The real capability map sits around advisor work.
The point is not to replace the CRM, planning tool, portfolio system, custodian, or compliance archive. The point is to help the firm prepare, compare, coordinate, and review the work that moves across them.
Values and risk capture
Organize goals, constraints, liquidity, tax context, values, risk profile, preferences, and open questions.
Reviewed client profile and source map.
IPS intelligence
Draft, update, compare, and flag IPS language against approved client inputs and portfolio policy.
Advisor-ready IPS redline and review log.
Proposal support
Assemble plan, portfolio, values, restrictions, assumptions, fees, and open issues before a proposal meeting.
Proposal context packet, not autonomous advice.
Portfolio-to-plan drift review
Compare current state against plan, IPS, stated values, risk profile, restrictions, and recent life events.
Review queue for advisor or investment team.
Review prep
Prepare meeting briefs, agenda suggestions, service history, open follow-ups, and client-specific questions.
Advisor-reviewed meeting packet.
Follow-up coordination
Turn meeting decisions into tasks, owner handoffs, drafts, and overdue-item review.
Clear next steps with owner and due date.
Compliance evidence support
Record workflow, data class, approved surface, reviewer, decision, and retained artifact.
Evidence of who reviewed what, when, and why.
Data-class method
Every AI decision starts with the data it touches.
A meeting-summary tool, research assistant, reporting workflow, and client-service copilot may all look similar in a demo. They are not similar once client-identifying, financial, sensitive, or privileged information enters the workflow.
A
Public
Risk tier
Public commentary, market education, published policy, website copy.
Safe for broad AI use when sources are checked.
B
Operational
Risk tier
Internal calendars, task lists, process notes, non-client project context.
Useful for workflow automation when access is scoped.
C
Internal confidential
Risk tier
Financials, strategy, employee issues, vendor contracts, management notes.
Use only in approved firm systems with logging and access control.
D
Client-identifying
Risk tier
Names, contact details, relationship notes, household references.
Needs an approved tool, limited access, and a clear business reason.
E
Client financial
Risk tier
Portfolio data, tax facts, balance sheets, income, holdings, planning inputs.
Needs advisor review, source-of-truth discipline, and strong vendor terms.
F
Client sensitive personal
Risk tier
Health, family, estate, employment, minor, identity, or vulnerability details.
Usually out of bounds unless counsel, compliance, and IT have agreed on controls.
G
Privileged / regulatory
Risk tier
Exam materials, counsel advice, complaints, supervision notes, privileged files.
Restricted. Do not use in everyday AI tools.
Decision matrix
Decide what AI can do before a workflow moves.
The matrix turns data classification into a rollout rule. It does not make legal promises. It gives the firm a way to decide what is open, what is controlled, and what stays out of bounds.
| Data | Internal use | Client work | Restriction |
|---|---|---|---|
| APublic | Internal use Open Drafting, summarizing, research support | Client work Open Education, web copy, market commentary | Restriction Controlled Review sources before publishing |
| BOperational | Internal use Open Calendar, task, meeting, and process help | Client work Controlled Limit to approved firm tools | Restriction Controlled Keep access role-based |
| CInternal confidential | Internal use Controlled Use only in controlled firm systems | Client work Controlled Logging and permissions required | Restriction No Not for open consumer tools |
| DClient-identifying | Internal use Controlled Use with clear business reason | Client work Controlled Advisor or staff review required | Restriction No No unmanaged tool entry |
| EClient financial | Internal use Controlled Only with strong vendor terms | Client work Controlled Systems of record stay authoritative | Restriction No No direct client output without review |
| FClient sensitive personal | Internal use No Usually out of bounds | Client work Controlled Exception only with counsel / IT input | Restriction No Never for casual experimentation |
| GPrivileged / regulatory | Internal use No Keep restricted | Client work Controlled Use only in explicitly approved context | Restriction No Never in general AI surfaces |
AI tool decision questions
Seven questions before a workflow moves from idea to rollout.
These are buying and rollout questions for the COO, CCO, IT provider, and business owner before a tool enters the team's routine.
1
What data class will this workflow touch?
2
Does the vendor use firm data to train shared models?
3
Is there a commercial agreement, data protection language, and firm control?
4
Can access be limited by role, workflow, and business need?
5
Can the firm see what happened later if someone asks?
6
Where does advisor review happen before anything reaches a client?
7
Have counsel, compliance, IT, and the business owner agreed on the rule?
Advisor review gates
AI can help prepare the work. It should not become the advisor.
Gate 1
AI prepares
Draft, summarize, compare, classify, extract, and surface useful context.
Gate 2
Person reviews
An advisor or qualified team member checks facts, tone, assumptions, and client fit.
Gate 3
Systems stay authoritative
Client records, planning tools, portfolio systems, and CRM remain the source of truth.
Gate 4
Exceptions get named
Sensitive personal details, privileged material, or new client-facing logic go to the right reviewer before rollout.
Operating environment map
AI should wrap around the people doing the work, not float outside the firm.
Good AI adoption is not a chatbot pasted onto the firm. It is a set of approved tools around employees, connected to the right systems, and governed by plain rules the firm can operate.
Frame
Regulatory and fiduciary obligations
The firm still carries the client promise. AI does not change that.
Fiduciary duty
Policies
Records
Counsel review
Work
Where employees already spend time
Email, meetings, documents, research, reporting, service requests, and follow-up.
Meetings
Inbox
Research
Reporting
AI
Approved assistants and workflow tools
The tools that can help only after the firm knows what data and review rules apply.
Drafting
Summaries
Extraction
Search
Operating rules
Use rules, review gates, access, and logging
The controls that turn AI from experimentation into approved routines.
Rules
Review
Access
Logging
Truth
Systems of record
Client records, planning tools, portfolio systems, CRM, and document stores stay authoritative.
CRM
Planning
Portfolio
Documents
Controls
Security, vendor, and oversight layer
Identity, permissions, retention, vendor terms, monitoring, and escalation paths.
Identity
DLP
Audit
Vendor terms
Who owns what
The work is clearer when every decision has an owner.
Firm
Business priorities, client promises, final risk appetite, advisor accountability, and the workflows worth changing.
Compliance counsel
Interpretation of obligations, policy language, supervisory expectations, and when a use case needs legal review.
IT provider
Identity, permissions, device posture, tenant settings, approved applications, security tooling, and implementation support.
Joans
Use-case selection, workflow design, tool decisions, rollout sequencing, training, and coordination across firm, counsel, and IT.
Readiness sequence
Move from scattered AI use to a usable first plan.
The sequence gives leadership a working map for what to use, what to control, what to defer, and who needs to review each next step.
Step 1
Find where AI is already showing up
Map where staff are already using AI, where partners see upside, and where client work is slow, repetitive, or inconsistent.
Step 2
Classify the data
Sort workflows by data class so the firm can separate safe experiments from work that needs tighter control.
Step 3
Choose the first tools and workflows
Decide which AI tools and workflows can move now, which need vendor or IT work, and which stay out of bounds.
Step 4
Set review gates and rollout path
Name the owner, advisor review step, training need, and next decision for each priority workflow.
Get the working version
Want the version we use to map this with your firm?
This public page is the field guide. The working version turns it into a firm-specific client journey map: AI opportunities, data exposure, advisor workflow support, review gates, clear owners, and a first rollout path.
No compliance promises. No generic AI tool list. Just the first working map for what your firm should move, control, and defer.
The working version is sent directly after we understand the AI work you are trying to move first.