AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Next Best Action

skill-joellewis-finance-skills-next-best-action · by JoelLewis

Design and implement next-best-action engines that surface proactive, prioritized recommendations to advisors based on portfolio, life, market, and compliance events. Use when the user asks about building event-driven advisor alerts, designing trigger logic for portfolio drift or large cash movements, prioritizing competing actions across a book of business, routing NBA recommendations to the rig…

No reviews yet
0 installs
3 views
0.0% view→install

Install

$ agentstack add skill-joellewis-finance-skills-next-best-action

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-joellewis-finance-skills-next-best-action)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
1mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Next Best Action? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Next-Best-Action — Event-Driven Advisor Recommendations

Core Concepts

Next-Best-Action Framework

Next-best-action is an advisor productivity and client service methodology that analyzes client data across systems to surface the single most valuable action an advisor should take for each client at any given time. The concept originates in CRM and marketing automation — industries that have long used event-driven recommendation engines to guide customer-facing personnel toward high-value interactions — but its application in wealth management addresses a distinct set of challenges: advisors managing hundreds of client relationships cannot manually monitor every portfolio, life event, compliance deadline, and practice touchpoint across their entire book of business.

NBA differs fundamentally from traditional task management. Traditional task management is reactive and manual: advisors create their own to-do lists, respond to inbound client requests, and rely on memory or periodic reviews to identify outreach opportunities. NBA is proactive and data-driven: the system continuously monitors client data across custodial feeds, CRM records, financial plans, compliance calendars, and market data, automatically identifying situations that warrant advisor attention and recommending specific actions with supporting context.

The core components of an NBA system are:

  • Event detection — Continuous monitoring of data sources to identify triggering events (portfolio drift, large cash movement, life milestone, compliance deadline, market dislocation).
  • Action identification — Mapping detected events to a catalog of recommended actions (schedule review, propose rebalancing, discuss tax-loss harvesting, update beneficiaries).
  • Prioritization — Scoring and ranking competing actions across all clients to ensure advisors focus on the highest-value activities given limited time.
  • Routing — Directing each action to the appropriate person based on role, expertise, relationship, and availability.
  • Delivery — Presenting recommendations through the channels advisors actually use (dashboard, mobile notification, email digest, CRM task).
  • Tracking — Recording action outcomes (accepted, deferred, rejected, completed) to close the feedback loop and improve future recommendations.

Event Detection and Trigger Types

The quality of an NBA system depends on the breadth and reliability of its event detection. Events fall into five categories, each requiring different data sources and detection logic.

Portfolio events are detected from custodial data feeds, portfolio management systems, and market data:

  • Drift beyond threshold — A client's actual allocation has deviated from the target model beyond the firm's tolerance band (e.g., equity allocation at 72% vs. 65% target with a 5% tolerance). Requires real-time or daily position data and model assignment.
  • Large cash deposit or withdrawal — A significant cash movement (typically defined by absolute amount or percentage of portfolio) has occurred. Detected from custodial transaction feeds. A $200,000 deposit into a $1 million account signals an investment opportunity; a $200,000 withdrawal from the same account may signal a liquidity event requiring plan reassessment.
  • Concentrated position — A single holding has grown to exceed a concentration threshold (e.g., 10% or 15% of portfolio value), whether through appreciation, additional purchases, or stock compensation vesting. Detected from position-level holdings data.
  • Tax-loss harvesting opportunity — Unrealized losses in taxable accounts exceed a significance threshold, particularly near year-end or after market declines. Requires lot-level cost basis data and market prices.
  • Required minimum distribution (RMD) due — A client with a traditional IRA or inherited IRA is approaching or has reached an RMD deadline. Requires account type data and client date of birth. RMD deadlines are absolute (December 31 for most, April 1 of the following year for the year the owner turns 73).
  • Margin call — A client's margin account has breached maintenance requirements. Requires margin balance and equity data from the custodian. Margin calls are time-sensitive and typically require same-day or next-day resolution.

Life events are detected from CRM data, client-reported information, and public records:

  • Birthday milestones — Age-based financial triggers: 59-1/2 (penalty-free IRA withdrawals), 62 (early Social Security eligibility), 65 (Medicare eligibility), 70-1/2 (qualified charitable distributions from IRAs), 73 (RMD beginning age under SECURE 2.0 for those born 1951-1959). Detected from client date of birth in CRM.
  • Marriage, divorce, death of spouse — Major life transitions requiring comprehensive financial plan review, beneficiary updates, account re-titling, and potentially revised investment strategy. Typically detected through advisor-reported CRM updates or client-initiated contact.
  • New child or grandchild — Triggers discussions about education savings (529 plans), life insurance review, estate plan updates, and beneficiary designation changes.
  • Job change or retirement — Income changes, employer benefit transitions (401k rollover), stock option/RSU vesting acceleration, and potential shift in investment time horizon and risk profile.

Market events are detected from market data feeds, portfolio analytics, and research systems:

  • Sector or asset class drawdown affecting client holdings — A significant decline in a sector or asset class in which the client has meaningful exposure. Requires mapping client holdings to sectors and monitoring sector-level returns.
  • Interest rate change impacting fixed income allocation — Significant rate movements that affect the duration risk, yield, or relative value of a client's fixed income holdings. Particularly relevant for clients with large bond allocations or approaching income-distribution phase.
  • New fund or product launch replacing a current holding — A lower-cost, better-performing, or more tax-efficient alternative to a fund currently held by clients. Detected through product research and comparison analytics.

Compliance events are detected from compliance calendars, CRM activity logs, and regulatory data:

  • Annual review overdue — The client has not received a formal portfolio or suitability review within the firm's required interval (typically 12 months). Detected by comparing the last review date in CRM to the current date.
  • Suitability or best-interest re-certification due — Client profile information is stale and requires re-confirmation. Particularly important under Reg BI, where the care obligation requires that recommendations reflect current client circumstances.
  • Disclosure delivery required — New regulations or rule amendments require delivery of updated disclosures (Form CRS updates, Form ADV amendments, privacy notices) to existing clients.

Practice events are detected from CRM activity tracking and practice management data:

  • Client contact gap — A client has not been contacted (by any channel) within the firm's service standard for that client's tier. For example, a Tier 1 client ($5M+ AUM) with a quarterly contact standard who has not been contacted in 100 days.
  • Upcoming contract renewal — An advisory agreement renewal or fee schedule review is approaching.
  • Referral opportunity — A client has recently had a positive experience (strong performance period, successful financial plan milestone, positive service interaction) that presents a natural referral conversation opportunity.

Action Library and Recommendation Logic

The action library is the catalog of all actions the NBA system can recommend. Each action is a defined, repeatable unit of advisor work with associated templates, context requirements, and completion criteria.

Action catalog design. A well-designed action library typically includes 30 to 60 distinct actions organized by category. Examples:

  • Schedule annual review meeting
  • Propose portfolio rebalancing to target model
  • Discuss tax-loss harvesting opportunity with estimated tax savings
  • Recommend Roth IRA conversion analysis (relevant for clients in temporarily low tax brackets)
  • Update beneficiary designations following life event
  • Review life insurance coverage adequacy
  • Discuss estate plan review (triggered by legislative change, asset growth, or family change)
  • Contact client regarding large uninvested cash position
  • Congratulate client on life milestone (birthday, retirement, grandchild)
  • Offer financial planning engagement to investment-only client
  • Present charitable giving strategy (donor-advised fund, qualified charitable distribution)
  • Discuss Social Security claiming strategy (approaching eligibility age)
  • Review held-away account for consolidation opportunity
  • Deliver required compliance disclosure

Action templates. Each action in the library includes pre-built supporting materials that reduce the advisor's preparation time and increase the likelihood of action completion:

  • Talking points — Key discussion topics tailored to the specific trigger and client context (e.g., "Your portfolio has drifted to 72% equities vs. your 65% target. I recommend we rebalance by trimming the overweight in large-cap growth and adding to international and fixed income. This trade would also harvest approximately $12,000 in losses to offset the gains we realized earlier this year.").
  • Email drafts — Pre-composed outreach emails personalized with client name, specific trigger details, and proposed next steps. The advisor reviews and edits before sending.
  • Meeting agendas — Structured agendas for review meetings, planning discussions, or specific topic conversations.
  • Analysis summaries — Pre-generated quantitative analysis (drift report, tax-loss harvesting estimate, RMD calculation, fee comparison) attached to the recommendation.

Recommendation logic. NBA systems use two primary approaches to map triggers to actions:

  • Rule-based logic — Deterministic IF-THEN rules that map specific triggers to specific actions. Example: IF client age reaches 72 AND has traditional IRA AND no RMD distribution recorded this year THEN recommend "Contact client re: RMD before December 31 deadline." Rule-based logic is transparent, auditable, and appropriate for compliance-driven and well-understood triggers.
  • ML-enhanced logic — Machine learning models that learn from historical advisor behavior to improve recommendations. The model observes which recommended actions advisors accept or reject, which actions lead to positive outcomes (client retention, additional assets, completed plans), and which client characteristics predict action relevance. ML enhancement is layered on top of rule-based logic — rules ensure required actions are never missed, while ML improves the ranking and presentation of discretionary actions.

Prioritization and Scoring

Not all actions are equal. An advisor with 200 clients might have 50 pending actions on any given day, but can realistically complete five to seven. Prioritization scoring determines which actions rise to the top of the queue.

Scoring dimensions. Effective prioritization considers multiple dimensions, each scored on a normalized scale:

  • Urgency — How time-sensitive is the action? An RMD deadline in December is more urgent in November than in March. A margin call is urgent immediately. A referral opportunity is low urgency. Urgency scoring should incorporate hard deadlines (compliance dates, regulatory deadlines) and soft deadlines (optimal timing windows that pass but do not create violations).
  • Impact — What is the potential benefit of completing the action? Impact can be measured along several sub-dimensions: revenue potential (will the action lead to new assets, a planning engagement, or retained AUM?), retention risk (is the client at risk of leaving if this issue is not addressed?), compliance requirement (is the action mandated by regulation or firm policy?), client satisfaction (will the action strengthen the relationship?).
  • Effort — How much advisor time and preparation does the action require? A quick congratulatory call requires five minutes; a comprehensive financial plan review requires several hours of preparation and a 90-minute meeting. Effort scoring ensures that the queue includes a realistic mix of quick wins and substantial engagements.
  • Client importance — What is the client's tier, AUM, relationship depth, and strategic value? Most firms weight actions for higher-tier clients more heavily, reflecting the disproportionate business impact of retaining and deepening large relationships. However, this weighting should not result in lower-tier clients being systematically neglected — compliance-driven and risk-driven actions should override tier-based scoring.

Composite scoring. The overall priority score is a weighted combination of dimension scores:

Priority = (Wurgency x Urgency) + (Wimpact x Impact) - (Weffort x Effort) + (Wclient x ClientImportance)

Weights are calibrated to the firm's strategic priorities. A firm focused on growth may weight impact and client importance heavily. A firm under regulatory scrutiny may weight urgency and compliance-driven impact heavily. Weights should be reviewed and adjusted periodically based on business outcomes.

Priority queue management. Queue design prevents action fatigue — the phenomenon where advisors ignore recommendations because the system generates too many:

  • Cap the daily queue at five to seven actions per advisor — a practical rule of thumb (not a cited research finding): larger queues tend to depress completion rates and breed action fatigue.
  • Ensure the queue includes a mix of action types — not exclusively compliance items or exclusively revenue-driven items.
  • Allow advisors to defer actions (with a snooze period) without permanently dismissing them. Deferred actions return to the queue after the snooze period with an adjusted priority score.
  • Compliance-driven actions with approaching deadlines should be flagged as non-deferrable once they enter a critical window (e.g., within 15 days of the deadline).

Advisor capacity consideration. The NBA system should account for advisor availability when generating queues. An advisor returning from vacation faces a backlog; an advisor with a light meeting day has capacity for a complex action. Integration with the advisor's calendar and CRM activity log enables capacity-aware queue generation.

Action Routing and Assignment

Not every action requires the advisor's personal attention. Effective routing directs each action to the right person based on the nature of the work, the relationship requirement, and the team structure.

Role-based routing rules:

  • Advisor — Relationship decisions, complex discussions, high-value client interactions, investment strategy conversations, financial planning reviews. The advisor handles actions where the personal relationship and professional judgment are essential.
  • Client service associate (CSA) — Scheduling meetings, document collection, routine follow-up calls, account maintenance requests, birthday and milestone acknowledgments (if the firm delegates these). CSAs handle actions that are important but do not require the advisor's expertise.
  • Operations team — Account transfers, beneficiary designation processing, cost basis corrections, distribution processing, account opening paperwork. Operations handles actions that involve back-office execution.
  • Compliance team — Regulatory disclosure delivery, suitability re-certification review, advertising review items, compliance-exception resolution. Compliance handles actions that require regulatory expertise or supervisory approval.

Escalation rules. Actions that are not completed within their service-level agreement (SLA) should escalate

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.