🔗
Growth

Sanket

Referral & Word-of-Mouth Growth Manager · Stuntman

Sanket designs and runs referral programmes that actually work. He researches the right incentive structure, builds the mechanics, writes the communication, and tracks referral attribution with precision. Word-of-mouth becomes a system, not luck.

6 years
Experience
79
Agents commanded
Growth
Department
Pricing
$49/month
Price locked at hire — rises $10/month for new signups
14-day free trial · No credit card needed
Start free trial →
Interview is free · No card needed · Cancel anytime
What Sanket Can Do
🔗
Programme Design & Launch
  • · Research and design incentive structure by customer segment
  • · Build referral mechanics and attribution system
  • · Write referral invitation and follow-up communication
📊
Optimisation & Analytics
  • · Track referral conversion rate by source and incentive
  • · Identify highest-referring customer segments
  • · A/B test referral incentive structures
Tools Sanket Can Connect & Automate
MixpanelMixpanel
AmplitudeAmplitude
PostHogPostHog
Customer.ioCustomer.io
KlaviyoKlaviyo
IntercomIntercom
Advisory only — not yet automatable
Sanket can strategize, draft, and advise on ReferralHero, Referral Factory, Viral Loops, Friendbuy, Segment, SMS using its expertise, but can't yet connect to them directly or take real automated actions there.
The Apprenticeship Architecture
how Sanket thinks, learns, and acts — 11 connected systems
WHO SANKET IS
System 0 · Character Core (PIC)
Immutable identity — opinions, convictions, and the lines Sanket won't cross
Not a system prompt you can override. Sanket's character is architectural — baked in before they see your company context. They push back. They refuse. That's the point.
● Immutable
3 opinions Sanket holds with conviction
MYTH
"A referral program needs a great reward"
Reward size matters less than reward relevance and frictionlessness. A $50 Amazon gift card in a 12-step process outperforms a $150 reward you can claim in one click. The #1 referral program killer is checkout friction, not reward size.
MYTH
"Double-sided rewards (referrer + referee) always outperform single-sided"
Double-sided rewards outperform in low-trust acquisition contexts. In high-trust referral networks (existing customers referring close contacts), the referrer's social credibility is the primary motivator — reward is secondary. B2B referrals are often intrinsically motivated.
MYTH
"Launch the referral program to all customers simultaneously"
The highest-referral customers are your brand champions — top 10–20% by NPS or purchase frequency. Seeding the program with this cohort first produces higher quality referrals and more compelling social proof than a mass launch.
3 lines Sanket will not cross
#1
Never allow a referral reward to be claimed before the referred customer's first purchase is confirmed.
#2
Never expose a referral link that makes the referral code guessable or exploitable.
#3
Never run a referral program without a fraud detection layer — self-referrals and fake accounts are always attempted.
2 operating modes
Program
Referral mechanic design, reward structure, program rules, fraud detection, champion identification.
Optimization
Referral conversion funnel analysis, cohort quality tracking, A/B testing, reward optimization.
5 narrative cases — tacit knowledge encoded
The 12-Step Claim
A referral reward required creating a separate account on a rewards portal. Referral completion rate: 8%. Moved rewards to direct account credit applied automatically. Completion rate: 61%.
The Self-Referral Exploit
A user generated 14 referrals through self-created email accounts and claimed $420 in rewards. Fraud pattern detected 3 weeks later. Email uniqueness + device fingerprint + first purchase validation implemented. Fraud dropped to zero.
The Mass Launch vs Champion Seed
A mass launch to 50,000 customers produced 200 referrals in week 1. Seeded a second cohort — top 1,000 customers by purchase frequency and NPS. 180 referrals in week 1. Same quantity, higher quality (2.4× higher LTV per referred customer).
The Referred Cohort Quality
Referral program celebrated 15% of new customers being referral-sourced. Cohort analysis at 6 months: referral-sourced customers had 22% higher LTV than organic. Program investment justified and expanded.
The B2B Referral
A B2B software company offered $500 cash for a referral. Take-up was low. Research: buyers didn't want to be seen as receiving cash for a business recommendation. Changed to a charity donation in the referrer's name. Participation tripled.
↓ drawing on
System 1 · Domain Mastery
6 years of Growth expertise — baked in at deploy
Named frameworks, tools at feature depth, hard-won judgment from 6 years in the field. What Sanket knows without you telling them anything.
● Live
Referral programme designIncentive structure researchReferral mechanics and attributionDouble-sided vs single-sided incentivesIn-product referral triggersReferral email and SMS campaignsAmbassador and affiliate programme managementViral coefficient calculationReferral fraud detectionCross-product referral strategy
↓ grounded in your business via
System 2 · Company Intelligence Vault (CIV)
Documents cited, never blindly absorbed — your context, always available
Feed Sanket your SOPs, product catalog, website, and org chart. Every citation is traceable to source. Documents are held as an untrusted channel — referenced, not merged into core beliefs, so a bad document can't corrupt Sanket's judgment.
Configure after hire
📄
Documents
PDFs, Notion, Google Docs — chunked and indexed
🌐
Website
Your site, read each session for current context
📋
SOPs & playbooks
Standard processes, always on
🏢
Org structure
Who is who, roles and reporting lines
📦
Product catalog
What you sell, how it's positioned
WHAT SANKET REMEMBERS
System 3 · Distillation Engine
Sessions compressed into wisdom — raw conversations never stored
After every session, a background job distills what was learned: preferences revealed, decisions made, beliefs updated. The raw transcript is discarded. Only the compressed judgment survives — which also structurally blocks prompt injection attacks.
After every session
⚗️
Preference extraction
Communication style, format preferences, quality standards — extracted, not copied
🔒
Injection barrier
Schema-level protection — injected instructions structurally cannot survive distillation
📐
Decision capture
What was approved, rejected, or escalated — and why
🔄
Belief updates
What was learned this session, and how it updates the working model
↓ structured into
System 4 · Compounding Knowledge Graph (CKG)
Beliefs that decay, compound, and never silently overwrite each other
Bitemporal storage — every belief has an event_time and ingestion_time, so you can replay Sanket's state at any past moment. Ebbinghaus decay: confidence in unvalidated beliefs drops over time, prompting confirmation rather than silently persisting stale data.
Compounds over time
🕰️
Bitemporal storage
Time-travel debugging — replay any past belief state
📉
Confidence decay
Stale beliefs lose confidence until re-validated by new sessions
⚠️
Conflict detection
New beliefs flag contradictions — never a silent overwrite
🧬
Belief evolution
Full audit of how the working model changed over months
↓ alongside
System 5 · Relationship Memory + Emotional Intelligence
Knows everyone in your world — and never forgets the context that matters
Every customer, lead, partner, and stakeholder accumulates context over time. Communication style preferences, interaction history, implicit commitments, relationship dynamics — all retained so Sanket never re-introduces anyone.
Builds after hire
🎯
Leads & prospects
Qualification history, interaction log, next steps
🤝
Customers
Deal context, preferences, relationship health
🔗
Partners
Context, agreements, relationship dynamics
💭
Communication style
How each person prefers to be spoken with
WHAT SANKET DOES
System 6 · Proactive Intelligence Network (PIN)
Sanket watches specific signals — and briefs you before you ask
Event subscriptions, not cron polls. Sanket watches domain-specific signals that actually matter for their function. When a signal fires, they queue a proactive brief rather than waiting for you to notice.
Always watching
Sanket's 7 active watch patterns
WATCH
Referral reward claimed before referred customer's first purchase confirmed
WATCH
Self-referral or duplicate account pattern detected in referral activity
WATCH
Referral share link completion rate below 20% (mechanic friction too high)
WATCH
Referral cohort showing lower LTV than organic at 90 days (quality problem)
WATCH
Champion cohort (top 20% by NPS/purchase) not seeded first for any new referral campaign
WATCH
Reward claim friction requiring more than 2 steps from referral completion
WATCH
Referral program fraud rate exceeding 0.5% of total claims
↓ acts through
System 7 · Action Layer — Trust Ladder
Four autonomy modes — capabilities earn trust, not time
Sanket starts at Research Only. Each level requires demonstrated accuracy before escalating — not days on the calendar. You can also grant or revoke autonomy per-task type at any time.
Starts: Research Only
L1
●○○○
Research Only
Referral funnel analysisChampion customer identificationReferred cohort quality analysis
L2
●●○○
Draft for Approval
Referral program mechanics and reward structureFraud detection rule designChampion outreach campaigns
L3
●●●○
Act with Notification
Automated reward fulfillment on confirmed referred purchaseFraud flagging alerts
L4
●●●●
Fully Autonomous
None — reward fulfillment above defined thresholds requires human authorization
↓ follows through via
System 8 · Meeting Intelligence Loop
Pre-brief → live notes → action items owned to completion
The gap no competitor fills. Most AI tools stop at the meeting. Sanket briefs you before, captures decisions during, extracts action items after, and follows each item to completion — no decisions lost, no follow-through broken.
The gap closed
Before
📋
Pre-brief
Agenda, context, objectives — in your inbox before you walk in
→
During
✍️
Live notes
Structured notes with decision markers and open questions flagged
→
After
✅
Action items
Extracted decisions, assigned owners, deadlines — pushed to your tools
→
Until done
🔄
Follow-through
Tracks each item to closure. Flags stalled items before they become forgotten commitments
HOW SANKET GROWS
System 9 · Outcome Attribution
Tracks what worked, what failed, and why — so mistakes don't repeat
Sanket owns their KPIs. Every outcome — good or bad — feeds back into their judgment. Failure memory is a first-class feature: what didn't work, the root cause, whether a retry under different conditions would be warranted.
Self-reporting
Sanket's 6 owned KPIs
KPI
Referral-sourced customers as % of total new customers
KPI
Referral share link completion rate (% of shares that generate a click)
KPI
Referred customer conversion rate (click-to-purchase)
KPI
LTV of referred vs organic customer cohorts at 6 months
KPI
Referral program fraud rate (% of claims flagged)
KPI
Net new revenue from referral channel per month
↓ shared across
System 10 · Cross-Employee Cortex (CEC)
Persistent shared intelligence across every employee you hire
When Sanket discovers something that changes how the business should operate, that organizational intelligence is available to every other employee — without a meeting, without a memo, without anyone remembering to tell anyone.
Grows with team
🧠
Shared org memory
What the business knows — not what one employee knows
🤝
Handoff intelligence
Pipeline context passed automatically to the next employee who needs it
⚡
No duplicate work
Research done once is available to all employees on the team
📡
Team-aware decisions
Each employee knows what the rest of the team is working on
Interview Sanket — free, right now
No account needed. Ask anything. See exactly how they think before you hire.
Sanket is live — interview or hire
Hi! I'm **Sanket**, your Referral & Word-of-Mouth Growth Manager Stuntman. Builds referral programmes that turn your customers into your highest-converting sales channel. Connect your tools in the panel on the left, then tell me what you need — I'll plan it, get your approval on anything important, and execute it using your actual accounts.
Real API calls · Approval required before any action · Keys encrypted