RevOps Agents

In Lessons 1-8, you used plugin commands one at a time -- /research for prospect intelligence, /score for lead qualification, /outreach for personalised messages. Each command produces a single output from a single input. Now you will build something fundamentally different: persistent, autonomous agents that manage revenue processes end-to-end.

The distinction matters. A /research command produces a brief. A Lead Intelligence Agent monitors your entire prospect list continuously, classifies signals in real time, updates scores automatically, and alerts the right rep within two hours of a buying window opening -- all without anyone typing a command. A /pipeline command produces a snapshot. A Revenue Reporting Agent aggregates pipeline data, marketing metrics, lead velocity, and forecast accuracy into a dashboard that lands in leadership inboxes every Monday morning -- automatically, with no manual effort.

This is the transformation that separates organisations using AI tools from organisations running AI-native revenue operations. The organisation that runs plugin commands has a better toolkit. The organisation that builds RevOps agents has a transformed revenue function.

What Is an Autonomous Agent?

An autonomous agent is a persistent AI workflow that runs on a schedule or in response to triggers, executes a defined sequence of steps, and produces outputs without requiring a human to initiate each step. Unlike a command (which you invoke), an agent runs itself. Unlike a chatbot (which waits for input), an agent actively monitors, analyses, and acts. The SKILL.md defines what the agent does, when it does it, and what it must never do.

What Is MCP (Model Context Protocol)?

MCP is the standard protocol that connects AI agents to external tools and data sources -- your CRM, email system, analytics platforms, calendar, and file storage. When we say an agent connects to your CRM "via MCP," we mean it uses a standardised connector that reads and writes CRM data programmatically. MCP connectors exist for Salesforce, HubSpot, Pipedrive, Gmail, Google Sheets, Slack, LinkedIn Campaign Manager, Google Analytics, and dozens of other platforms. The agent does not need separate API integrations for each tool -- MCP provides a universal interface.

The Five Core RevOps Agents

Agent	Purpose	Schedule	Key Integration
Lead Intelligence	Monitor prospects; alert on buying signals	Daily scan	Web Search, LinkedIn, CRM, Google News
CRM Hygiene	Keep data current; flag stale records	Weekly + Monthly	CRM, LinkedIn, Email Verification
Outreach Sequencing	Manage multi-touch sequences; track engagement	Continuous	CRM, Gmail/Outlook, LinkedIn
Marketing Performance	Pull channel data; generate weekly analysis	Every Friday	LinkedIn Ads, GA4, Email Platform, CRM
Revenue Reporting	Produce leadership dashboard with pipeline and forecast	Every Monday	CRM, Marketing Platform, Finance System

Agent 1: Lead Intelligence Agent

Purpose: Monitor configured prospect lists and signal sources continuously. Alert sales reps within two hours when HOT timing signals appear. Never let a buying window go undetected.

The Lead Intelligence Agent is the front line of your revenue engine. Without it, buying signals -- a prospect announcing a funding round, a target contact posting about the exact problem you solve, a company posting four operations roles in a week -- pass unnoticed. Your competitors' reps reach out first. The window closes.

Architecture: The agent connects via MCP to web search (Google News, trade press), LinkedIn (company pages, contact activity), Companies House or equivalent registries (filings, director changes), and your CRM (account assignments, current scores). It scans daily against your configured prospect list.

Signal Classification:

The agent classifies every detected event into three tiers:

Tier	Alert Timing	Criteria	Examples
HOT	Within 2 hours	At least 2 corroborating data points; clear buying signal	Funding round announced; major contract win; new VP hire in target role; pricing page visit
WARM	Daily digest	Single positive signal; growth indicator	New job posting in target department; expansion announcement; conference speaking slot
NEUTRAL	Weekly digest only	General industry context; no specific prospect relevance	Industry news; competitor announcements (shared with leadership, not reps)

Trigger Conditions: The agent fires a HOT alert when a monitored account shows a combination of signals that indicate active buying readiness. A single data point -- one job posting, one news article -- is never enough for HOT classification. The agent requires corroboration: a funding round plus rapid hiring, or a new contract win plus a target contact posting about operational scaling pressure.

Failure Modes: The most common failure is alert fatigue -- too many WARM signals promoted to HOT, causing reps to ignore alerts. The classification thresholds must be calibrated against your historical conversion data: which signals actually preceded your last ten closed-won deals? Those are your HOT criteria. Everything else is WARM or NEUTRAL.

Sample Weekly Signal Report:

LEAD INTELLIGENCE DIGEST -- Week of 10 March 2026
================================================================

HOT signals since last digest:      3 -- see separate alerts below
WARM signals:                       11
Accounts monitored:                 127

HOT SIGNAL ALERTS

  (1) Meridian Logistics -- Leeds
      Signal:    Series B funding (GBP 18M) announced via TechCrunch, 7 Mar
                 + 4 Operations roles posted in last 14 days
      Contact:   Sarah Chen, VP Operations
      Rep:       James Wright (North territory)
      Score:     Updated from 54 --> 87
      Action:    Run /research + /outreach immediately
      Deadline:  First touch by 12 March

  (2) Vantage Distribution -- Birmingham
      Signal:    New CEO appointed (ex-DHL, started 1 Feb)
                 + Company website redesigned with "digital transformation" messaging
      Contact:   Mark Evans, COO
      Rep:       Priya Sharma (Midlands territory)
      Score:     Updated from 41 --> 72
      Action:    Run /research; CEO change = reset relationship; new outreach angle

  (3) Horizon Fulfilment -- Manchester
      Signal:    Pricing page visit (9 Mar, 14:22)
                 + Downloaded "Scaling Ops Without Headcount" whitepaper (same session)
      Contact:   Rachel Adams, Director of Operations
      Rep:       Tom Bailey (North-West territory)
      Score:     Updated from 63 --> 78
      Action:    Deploy Touch 1 within 24 hours; lead with whitepaper content

WARM SIGNALS

  Apex Warehousing     -- New depot lease filed (Nottingham Council, 6 Mar)
  BrightPath Logistics -- CFO posted about margin pressure (LinkedIn, 8 Mar)
  Cornerstone 3PL      -- Job posting: "Process Improvement Manager" (Indeed, 5 Mar)
  [... 8 additional entries ...]

SCORE CHANGES (crossed 60 threshold this week)
  Vantage Distribution:  41 --> 72 (HOT)
  BrightPath Logistics:  55 --> 64 (newly HOT)

ROLE CHANGES DETECTED
  David Nguyen -- was Head of Ops at Pinnacle Logistics; now VP Supply Chain
  at Clearwater Distribution. Action: Update CRM; reassign to South territory;
  new research required.

================================================================

This report is generated automatically every Monday morning and distributed to RevOps and sales leadership. Each HOT alert also triggers a separate immediate notification to the assigned rep within two hours of detection.

Agent 2: CRM Hygiene Agent

Purpose: Keep CRM data current automatically. Run enrichment on a defined schedule. Flag stale, incorrect, or incomplete records. Ensure every rep works with accurate, complete prospect data at all times.

CRM data decays at approximately 30% per year. People change jobs. Companies pivot. Email addresses bounce. Phone numbers change. A record entered eighteen months ago may be partially or entirely wrong today -- and most records were never complete to begin with. The CRM Hygiene Agent eliminates data decay as a revenue risk.

What Is an SLA (Service Level Agreement)?

In the context of RevOps agents, an SLA is not a customer contract -- it is an internal commitment defining how quickly the team must act on a signal. When we say "Tier 1 SLA: 1 business day," we mean the assigned rep must respond to a Tier 1 lead within one business day. SLA tracking ensures that leads do not sit unworked while buying signals decay.

Architecture: The agent connects via MCP to your CRM (Salesforce, HubSpot, Pipedrive), web search (for real-time firmographic updates), LinkedIn (for role verification and activity monitoring), and email verification APIs (for deliverability checks). It runs three scheduled workflows at different cadences.

Scheduled Workflows:

Cadence	Scope	Actions
Weekly	All HOT and WARM scored records with last enrichment older than 7 days	Verify email deliverability; check for role changes; scan for new timing signals; update scores; alert Lead Intelligence Agent if any score crosses 60 threshold
Monthly	All Tier 1 accounts plus full active pipeline	Full enrichment pass; revenue and headcount refresh; tech stack update via job posting analysis; score recalculation; identify contacts who changed roles; flag significant score declines to customer success team
Quarterly	All active records in CRM	Full database enrichment; email bounce cleanup and re-verification; duplicate contact detection and merge recommendations; inactive record archiving (no activity for 12+ months and score below 20); comprehensive data quality report

Data Quality Scoring: The agent maintains a data quality score for the entire CRM, measured as the percentage of records that are complete (all mandatory fields populated, email verified, last enrichment within 90 days). This score trends over time, giving RevOps leadership visibility into whether data quality is improving, stable, or declining.

Deduplication Logic: The quarterly pass includes duplicate detection using fuzzy matching on company name, contact email, and phone number. The agent does not merge duplicates automatically -- it produces a merge recommendation report for RevOps to review and approve. Automated merges risk losing data; human review ensures the surviving record retains all relevant history.

Failure Modes: The most dangerous failure is silent data decay -- records that look current but contain stale information. The agent mitigates this by tracking "last enrichment date" for every record and escalating any record where enrichment has not run within the configured window. The second failure mode is over-enrichment -- running expensive verification on records that do not warrant it. The tiered scheduling (weekly for HOT/WARM, monthly for Tier 1, quarterly for all) ensures enrichment effort is proportional to record value.

Sample Hygiene Report:

CRM HYGIENE REPORT -- Monthly | 1 March 2026
================================================================
Records processed:     1,247
Records updated:       389 (31.2%)
Records unchanged:     858 (68.8%)

CHANGES MADE
  Role changes:        23 contacts updated (new title or employer)
  Email corrections:   14 addresses updated (bounced --> verified replacement)
  Revenue updates:     31 accounts (Companies House annual filings)
  Headcount changes:   47 accounts (LinkedIn company page delta)
  New timing signals:  8 accounts flagged (see HOT signals below)

FIELD UPDATE FREQUENCY (top 5 most-changed fields)
  1. Contact title         -- 23 records
  2. Account headcount     -- 47 records
  3. Contact email         -- 14 records
  4. Account revenue est.  -- 31 records
  5. Tech stack indicators -- 19 records

ROLE CHANGES REQUIRING MANUAL ACTION
  David Nguyen -- was Head of Ops at Pinnacle Logistics; now VP Supply Chain
  at Clearwater Distribution. Action: Re-route to South territory rep.
  Lisa Park -- was Director of Logistics at Apex Warehousing; now COO at
  Summit Freight (new company, not in CRM). Action: Create new account record.

NEW HOT SIGNALS IDENTIFIED DURING ENRICHMENT
  BrightPath Logistics -- Posted 3 warehouse roles in 2 weeks; score 55 --> 64
  Clearwater Distribution -- New VP hire (David Nguyen, ex-Pinnacle); score 38 --> 58

DATA QUALITY ISSUES
  Missing email:          34 records
  Bouncing email:         14 records (corrected 14; 0 remaining unresolved)
  Missing phone:          89 records
  Last enrichment >90d:   112 records -- recommend: schedule ad-hoc enrichment pass

DATA QUALITY SCORE:     78.4% of records complete
Prior period:           74.1%
Trend:                  Improving (+4.3 percentage points)
================================================================

Agent 3: Outreach Sequencing Agent

Purpose: Manage multi-touch outreach sequences for all HOT and WARM leads. Track which touches have been sent. Trigger follow-ups on schedule. Flag replies immediately for rep attention. Never miss a touch. Never continue a sequence after a prospect has replied.

The Outreach Sequencing Agent turns the one-time outputs of the /sequence command into a managed, tracked, continuously executing process. Without it, sequences are generated but not managed -- reps forget to send Touch 3, miss the timing window for Touch 5, or continue sending automated messages after the prospect has already replied (the single most damaging outreach error in B2B sales).

Architecture: The agent connects via MCP to your CRM (for sequence status tracking), email platform (Gmail or Outlook, for send, open, click, and reply detection), and LinkedIn (where available, for connection and message tracking). It runs continuously, checking for pending touches and engagement signals.

Sequence Lifecycle:

The agent manages four distinct phases of every sequence:

1. Initiation -- When a lead is classified HOT, the agent automatically triggers /research to generate a brief, then /sequence to build a multi-touch sequence, then schedules Touch 1 within 24 hours of HOT classification. Status in CRM: ACTIVE.

2. Execution -- The agent sends each touch at the scheduled time, logs the send in CRM with timestamp and channel, and checks for engagement (opens, clicks, replies) before sending the next touch. Default HOT timing: Day 1, 3, 7, 10, 17, 21. Default WARM timing: Day 1, 7, 14, 28, 35.

3. Branching -- The agent monitors for three events that alter the sequence flow:

Event	Agent Action
Reply received	Immediately pause sequence (status: REPLIED); alert assigned rep within 15 minutes; log in CRM; do NOT send any further automated touches
Bounce detected	Pause sequence (status: PAUSED); trigger /enrich to find updated email; alert rep: "Email bounced -- verify contact before resuming"
Unsubscribe received	Stop all touches on all channels permanently (status: OPTED-OUT); update CRM: DO NOT CONTACT via email; alert rep for potential phone or LinkedIn follow-up

4. Completion -- If all touches are sent with no response, the agent logs "Sequence completed -- no response" in CRM, moves the prospect to a monthly CULTIVATE nurture cadence managed by marketing, and schedules a re-evaluation via /enrich and /score in 90 days.

Timing Optimisation: The agent adjusts timing based on engagement signals. If a prospect opens an email but does not click, the agent brings the next touch forward by three days (engagement signal suggests interest but insufficient motivation -- a faster follow-up catches the window). If a prospect opens multiple emails without any click, the agent does not send the final "closing the loop" touch -- instead, it substitutes one additional value-add touch (a case study or data point), because consistent opens indicate the prospect is reading and may need a stronger hook rather than a graceful exit.

Failure Modes: The critical failure is sending a touch after a reply. Any reply -- even "not interested right now" -- means the conversation is now human-led. The agent's NEVER rules enforce this absolutely. The second failure is sending too many touches on the same channel within a short window, which triggers spam filters and damages sender reputation. The minimum gap rule (3 days between touches on the same channel) prevents this.

Sample Sequence Status Output:

SEQUENCE STATUS -- Active Sequences Summary | 10 March 2026
================================================================
Active sequences:      34
Replied (awaiting rep): 7
Completed (no reply):  12
Opted-out:              2
Paused (bounce):        1

TOP PRIORITY -- REPLIES AWAITING REP ACTION

  Sarah Chen / Meridian Logistics
  Sequence: HOT-6 | Touch 2 (email) | Replied 9 Mar 16:42
  Reply: "Thanks for the note. We are actually looking at this right now.
         Can you send me some more detail on what you do for 3PLs?"
  Rep: James Wright | Alert sent: 9 Mar 16:43 | Rep acknowledged: 9 Mar 17:15
  Status: REPLIED -- sequence paused; rep leading conversation
  Recommended: Run /brief type:"pre-call" and propose a discovery call

  Mark Evans / Vantage Distribution
  Sequence: HOT-6 | Touch 1 (LinkedIn DM) | Replied 8 Mar 09:20
  Reply: "Interesting timing -- let's talk. Next week?"
  Rep: Priya Sharma | Alert sent: 8 Mar 09:21 | Rep acknowledged: 8 Mar 09:35
  Status: REPLIED -- discovery call being scheduled

UPCOMING TOUCHES (next 48 hours)
  Rachel Adams / Horizon Fulfilment -- Touch 3 (LinkedIn value-add) -- 11 Mar
  Tom Jennings / Apex Warehousing -- Touch 4 (email reframe) -- 11 Mar
  Claire Watson / BrightPath -- Touch 1 (LinkedIn DM) -- 12 Mar

BOUNCE REQUIRING ACTION
  Mike Torres / Cornerstone 3PL -- Touch 2 bounced (email invalid)
  Enrichment triggered: awaiting updated email address
  Rep alert sent: Tom Bailey
================================================================

Agent 4: Marketing Performance Agent

Purpose: Automatically pull weekly performance data from all connected marketing channels. Generate analysis with the top three optimisation recommendations. Distribute to relevant stakeholders. Track progress against campaign KPIs.

Most marketing teams review performance manually -- someone pulls data from LinkedIn Ads, someone else checks Google Analytics, a third person runs an email report, and the marketing manager spends two hours assembling a narrative. The Marketing Performance Agent eliminates this manual assembly entirely. Every Friday, it pulls data from every connected channel, compares actuals to targets, identifies trends, and produces an analysis report with specific, actionable recommendations -- not just observations.

What Is Anomaly Detection?

In the context of the Marketing Performance Agent, anomaly detection is the automatic identification of metrics that deviate significantly from expected values. If your landing page conversion rate drops 35% week-over-week, the agent does not wait until the Friday report to flag this -- it sends an immediate alert. Anomaly detection catches problems early, before they compound into missed targets. The agent uses configurable thresholds: you define what "significant deviation" means for each metric based on your historical data.

Architecture: The agent connects via MCP to LinkedIn Campaign Manager (impressions, clicks, CTR, CPL), Google Analytics GA4 (sessions, conversion rates, engagement), your email platform (HubSpot, Mailchimp, or Klaviyo -- send volume, open rates, click rates, unsubscribe rates), your CRM (leads by source, score distribution, pipeline attribution), and ad platforms (Google Ads, Meta Ads if applicable).

Automated Schedule: Every Friday (or configured day), the agent executes a five-step workflow:

1. Pull data from all connected channels via MCP (preceding 7 days)
2. Compare actuals to targets from the campaign brief in local configuration
3. Compare actuals to prior week for trend analysis
4. Run the /analyze workflow to identify the top 3 optimisation opportunities
5. Generate the weekly performance report and distribute it

Anomaly Alert Triggers: The agent sends immediate alerts (not waiting for the weekly report) when any of these thresholds are breached:

• Email unsubscribe rate exceeds 0.5% in any single send
• Ad spend is depleted before the end of the scheduled period
• Landing page conversion drops more than 30% week-over-week
• HOT lead volume drops more than 25% week-over-week with no explained cause
• CRM integration error detected: leads not flowing from marketing to sales

Failure Modes: The primary failure mode is false positives -- alerting on normal variance. A single bad email send with a 0.6% unsubscribe rate may not indicate a systemic problem. The agent mitigates this by requiring sustained anomalies (two consecutive weeks or a single-week deviation exceeding the threshold by 50%) before escalating to leadership. The secondary failure mode is stale data connections -- if an MCP connector fails silently, the report will show incomplete data. The agent validates data completeness before generating the report and flags any channel with missing or incomplete data.

Sample Weekly Performance Dashboard:

MARKETING PERFORMANCE REPORT -- Week of 3 March 2026
Campaign: Q2 Lead Gen -- "The Scaling Problem"
================================================================

HEADLINE NUMBERS -- WEEK 5 OF 12

                        Actual    Target    vs Target    vs Last Week
Whitepaper downloads:      47        35     134%  (+)       +12%
HOT leads generated:        6         4     150%  (+)        +2
LinkedIn impressions:  18,400    12,000     153%  (+)        +8%
LinkedIn CTR:           0.38%     0.40%      95%  (-)       flat
Email open rate:          34%       28%     121%  (+)        +2%
Email click rate:        3.1%      4.2%      74%  (!)       -0.3%
Cost per download:      GBP 8.50  GBP 12   141%  (+)       -GBP 1.20
Cost per HOT lead:    GBP 381   GBP 500    131%  (+)      -GBP 44

TOP 3 OPTIMISATION OPPORTUNITIES

1. EMAIL CLICK RATE IS DRAGGING (3.1% vs 4.2% benchmark)
   Open rate is strong (34%) -- subject lines work. Click rate is weak.
   Diagnosis: Email 3 in nurture sequence ("3 patterns we see in 3PL ops")
   has lowest click rate at 1.8%. It is too theoretical -- lacks a concrete hook.
   RECOMMENDED: Rewrite Email 3 with a specific case study. Test subject line:
   "How one Midlands 3PL doubled throughput without hiring."
   Expected impact: +1-1.5% click rate on Email 3; +3-4 incremental HOT leads.

2. THOUGHT LEADER ADS OUTPERFORMING SPONSORED CONTENT (2.8x CTR)
   CEO personal profile posts: 1.06% CTR. Company sponsored content: 0.38% CTR.
   Current budget split: 70% Sponsored / 30% Thought Leader Ads.
   RECOMMENDED: Rebalance to 40% Sponsored / 60% Thought Leader Ads for weeks
   6-12. Reallocate GBP 2,500.
   Expected impact: +35% total LinkedIn CTR; +8-10 additional downloads/week.

3. TRADE PRESS CONVERSION BELOW EXPECTATION
   Motor Transport article (Week 4): 31 UTM-tracked visits, only 8 downloads
   (26% conversion vs 58% from other sources). Trade press audience is in
   awareness mode -- whitepaper landing page is too mid-funnel for them.
   RECOMMENDED: Create a lighter entry point ("3PL Scaling Checklist," 1 page)
   as the trade press landing page, with whitepaper as upsell after email capture.
   Expected impact: +15-20% conversion from trade press traffic.

WHAT IS WORKING -- DO NOT CHANGE
  LinkedIn audience targeting: 94% of HOT leads within target firmographics
  Whitepaper engagement: 73% of downloaders reading past page 6
  Sales Plugin integration: average enrichment within 18 hours of download;
  average HOT lead to first outreach within 31 hours

================================================================

Agent 5: Revenue Reporting Agent

Purpose: Produce the weekly revenue dashboard for sales and marketing leadership. Pipeline by stage, lead velocity, campaign contribution, forecast vs. target. Every Monday morning -- automated, no manual effort.

The Revenue Reporting Agent is the single view of truth for the revenue function. Without it, the VP of Sales asks three people for pipeline numbers, the CMO assembles marketing contribution data manually, and the CEO gets a different answer depending on who they ask and when. The agent eliminates this by aggregating all data sources into a single, consistent, automatically generated dashboard.

Architecture: The agent connects via MCP to your CRM (pipeline data, deal stages, close dates, rep assignments), marketing platform (lead source attribution, campaign spend), and finance system if available (actual revenue vs. forecast). It runs every Monday morning (or configured day) and distributes the dashboard to configured recipients.

Dashboard Metrics: The agent produces four metric categories:

Pipeline Metrics -- Total pipeline value (all stages); weighted pipeline (value multiplied by stage probability); pipeline coverage ratio (total pipeline divided by quarterly quota remaining); pipeline created this week vs. target; average deal size vs. target; average sales cycle length vs. target; pipeline at risk (deals with no activity for more than 14 days, with count and currency value); deals closed this week (won, lost, and no-decision).

Lead Velocity Metrics -- HOT leads generated this week vs. target; MQL to SAL conversion rate (this week vs. four-week rolling average); SAL to Opportunity conversion rate; time from HOT classification to first outreach (average, this week); time from Opportunity to Close (average, all deals closed this period).

Marketing Contribution -- Pipeline generated from marketing-sourced leads (currency value); percentage of total pipeline attributed to marketing; top three performing lead sources this week (by HOT lead volume); campaign spend this week vs. budget.

Forecast -- Committed this quarter (deals with greater than 80% probability): currency value and deal count; likely this quarter (50-80% probability): currency value and deal count; upside (below 50% probability): currency value and deal count; current forecast vs. quarterly target as a percentage.

Leading Indicator Alert System: The agent monitors one critical leading indicator: the MQL-to-SAL conversion rate. If this rate drops below the configured threshold in any given week, the agent sends an immediate alert to RevOps and VP Sales with the message: "Revenue risk in 60-90 days." This is the earliest signal that the revenue engine is slowing -- it appears two full quarters before it shows up in closed-won revenue.

Failure Modes: The primary failure is data lag -- if CRM data is not updated by reps, the dashboard shows stale pipeline. The agent mitigates this by flagging "deals with no activity for more than 14 days" prominently in the at-risk section, creating accountability pressure. The second failure mode is forecast inflation -- reps overstate deal probability. The agent addresses this by showing weighted pipeline alongside raw pipeline, and by tracking forecast accuracy over time (what percentage of "committed" deals actually closed in prior quarters).

Sample Monday Morning Revenue Dashboard:

REVENUE DASHBOARD -- Week of 10 March 2026
Prepared: Monday 06:00 | Distributed to: VP Sales, CMO, CEO, RevOps
================================================================

PIPELINE METRICS
                              Actual        Target      Status
Total pipeline:           GBP 2.4M          --           --
Weighted pipeline:        GBP 1.1M          --           --
Coverage ratio:              3.2x          3.0x          OK
Pipeline created (week):  GBP 185K      GBP 150K        (+)
Avg deal size:            GBP 42K       GBP 45K         (-)
Avg cycle length:            38 days       35 days       (-)
Pipeline at risk:         GBP 340K (6 deals, no activity >14 days)

LEAD VELOCITY
HOT leads this week:          8              6            (+)
MQL to SAL conversion:       38%            35%           (+)
  (4-week average:           36%)
SAL to Opportunity:          62%            60%           (+)
HOT to first outreach:       22 hours       24 hours      (+)
Opportunity to Close:        34 days        35 days       (+)

MARKETING CONTRIBUTION
Pipeline from marketing:  GBP 95K    (51% of total pipeline created this week)
Top sources: LinkedIn Sponsored (3 HOT), Whitepaper organic (2 HOT), Referral (1)
Campaign spend (week):    GBP 2,100  of GBP 2,400 budget  (87% -- on track)

FORECAST -- Q1 2026
Committed (>80%):        GBP 480K     7 deals    (target: GBP 600K)
Likely (50-80%):         GBP 310K     5 deals
Upside (<50%):           GBP 220K     4 deals
Forecast vs target:          80%                  -- needs 3 more commits

THIS WEEK'S HIGHLIGHTS
  WIN:  Meridian Logistics replied to outreach; discovery call booked for Thursday
  RISK: 6 deals (GBP 340K) with no CRM activity >14 days -- rep action needed
  ASK:  3 committed deals need executive sponsor call to close this quarter
        -- VP Sales to schedule

AT-RISK DEALS REQUIRING LEADERSHIP ATTENTION
  Summit Tech    | J. Wright  | GBP 65K  | No activity 21 days | Recommend: call
  Vanguard Ltd   | P. Sharma  | GBP 48K  | Competitor engaged  | Recommend: exec mtg
  Northern Ops   | T. Bailey  | GBP 72K  | Budget delayed       | Recommend: Q2 reforecast

================================================================

Agent Orchestration: How the Five Agents Work Together

The five agents are not independent -- they form an orchestrated system where each agent's output can trigger another agent's workflow.

                    SIGNAL SOURCES
                    (Web, LinkedIn, News, CRM)
                           |
                           v
              +---------------------------+
              |  Lead Intelligence Agent  |  Daily scan
              |  (monitors, classifies,   |  HOT alerts within 2 hours
              |   updates scores)         |
              +---------------------------+
                    |              |
           HOT alert         Score update
                    |              |
                    v              v
    +-------------------+   +-------------------+
    | Outreach Sequencing|  | CRM Hygiene Agent |  Weekly + Monthly
    | Agent              |  | (enrichment,      |  enrichment passes
    | (builds sequence,  |  |  verification,    |
    |  manages touches,  |  |  deduplication)   |
    |  tracks replies)   |  +-------------------+
    +-------------------+         |
           |                Score crosses 60
           |                      |
           v                      v
    Rep gets alert    Lead Intelligence Agent
    (reply detected)  (generates new HOT alert)

    +----------------------------+
    | Marketing Performance Agent|  Every Friday
    | (channel data, analysis,   |
    |  optimisation recs)        |
    +----------------------------+
              |
              v
    +----------------------------+
    | Revenue Reporting Agent    |  Every Monday
    | (pipeline, velocity,       |
    |  forecast, dashboard)      |
    +----------------------------+
              |
              v
        Leadership inbox

Scheduling Patterns:

Day	Agent Activity
Daily	Lead Intelligence Agent scans all signal sources; Outreach Sequencing Agent sends scheduled touches and checks for replies
Friday	Marketing Performance Agent pulls weekly channel data and generates analysis report
Monday	Revenue Reporting Agent aggregates pipeline, velocity, and forecast data into leadership dashboard
Weekly	CRM Hygiene Agent runs enrichment on HOT and WARM records
Monthly	CRM Hygiene Agent runs full enrichment pass on all Tier 1 accounts; Lead Intelligence Agent adds newly ICP-matched companies to watch list
Quarterly	CRM Hygiene Agent runs full database enrichment, bounce cleanup, and duplicate detection

Inter-Agent Handoffs: The critical handoff is between the CRM Hygiene Agent and the Lead Intelligence Agent. When the CRM Hygiene Agent's monthly enrichment pass discovers a new timing signal (a company that was WARM now shows funding plus hiring), it updates the score. If the score crosses the 60-point threshold, the Lead Intelligence Agent automatically fires a HOT alert to the assigned rep. This chain -- enrichment discovers signal, score updates, alert fires, sequence initiates -- runs without any human triggering any step.

Try With AI

Use these prompts in Claude or your preferred AI assistant to explore this lesson's concepts.

Prompt 1: Design a Lead Intelligence Agent

I am building a Lead Intelligence Agent for my organisation.
My ICP is [describe your ideal customer -- industry, company size,
target persona]. My product solves [describe the core problem].

Design the complete agent SKILL.md:
1. Signal sources to monitor (primary and secondary)
2. Signal classification rules (HOT / WARM / NEUTRAL) with
   specific examples for my industry
3. Alert format with recommended actions
4. Daily digest format
5. The NEVER rules (what the agent must not do)
6. How this agent hands off to the Outreach Sequencing Agent
   when a HOT alert fires

Present as a deployable SKILL.md file.

What you are learning: Designing a monitoring agent requires thinking about signal quality, not just signal quantity. The classification rules and the corroboration requirement (two data points for HOT) are the governance layer that prevents alert fatigue. The NEVER rules define what the agent must not do autonomously -- these boundaries are as important as the workflows.

Prompt 2: Map the Agent Orchestration

I have these five RevOps agents: Lead Intelligence, CRM Hygiene,
Outreach Sequencing, Marketing Performance, Revenue Reporting.

Walk me through this scenario step by step, identifying which
agent is responsible at each stage:

1. It is Tuesday. The CRM Hygiene Agent runs its weekly enrichment
   pass and discovers that Apex Corp (currently scored 48, WARM)
   has just announced a Series A funding round and posted 3 new
   engineering roles.

2. The score is recalculated. What happens next?

3. By Thursday, the prospect replies to Touch 1. What happens?

4. On Friday, the marketing report runs. Where does this prospect
   appear?

5. On Monday, the revenue dashboard runs. Where does this deal
   appear?

For each step, name the agent, the trigger, and the output.

What you are learning: Revenue operations agents do not work in isolation -- they form a system. Tracing a single prospect through all five agents reveals the inter-agent handoffs and shows how a buying signal detected on Tuesday becomes a pipeline entry on Monday's dashboard with no manual data entry at any step.

Prompt 3: Build for Your Market

I operate in [your market -- e.g., B2B SaaS in South Asia,
professional services in the GCC, fintech in Southeast Asia].

For this market, what signal sources should my Lead Intelligence
Agent monitor? Consider:
- Local business registries (equivalent to Companies House)
- Regional trade press and industry publications
- Local social media platforms (not just LinkedIn)
- Regional funding databases
- Government procurement portals (if relevant)
- Local job boards

Also consider: what cultural factors affect outreach timing and
channel selection in this market? How should my Outreach Sequencing
Agent adapt its channel mix and timing for this region?

What you are learning: The RevOps agent architecture is universal, but the signal sources, timing, and channel preferences are market-specific. An agent monitoring the Pakistan market should track SECP filings, Dawn Business, and local job portals alongside LinkedIn. An agent in the GCC should monitor DIFC/ADGM registrations, Gulf News Business, and WhatsApp as a primary outreach channel alongside email.

---

Continue to Lesson 10: The Skill Library ->