We build revenue-generating infrastructure for category leaders.

Phase 1: Foundation Setup (Weeks 1-2)

1.1 Customer Avatar Development

Before building anything, we needed to know who we were building for. The marketing automation specialist went on to conduct a full quantitative and qualitative analysis of Rootwell's existing customer base. The results we received:

• Quantitative analysis involved pulling the last 12 months of Shopify transaction data, Google Analytics behaviour data, and email engagement metrics. We built pivot tables segmenting customers by age range, geographic distribution, purchase frequency, average order value, and product preference. For a brand like Rootwell, we expected to find 3–4 distinct buyer clusters.

• Qualitative research involved interviewing 10–15 existing customers across segments — using a 15-question interview framework covering pain points, decision-making process, objections, preferred communication channels, and purchase triggers. The interviews revealed the why behind the quantitative patterns.

Projected personas for a brand like Rootwell:

• The Overwhelmed Professional (35 - 48, urban, $75K+ income): High stress load, research-driven, sceptical of marketing claims, values scientific evidence, purchases ashwagandha and focus formulas. Buys based on clinical study citations and peer recommendations. Discovery channel: podcasts and LinkedIn.

• The Wellness Optimiser (26 - 35, health-conscious, active lifestyle) Already takes supplements, looking to optimise their stack, highly engaged on Instagram, responds to educational content about mechanisms of action. Buys the daily wellness stack. Discovery channel: Instagram and wellness blogs .

• The Curious Newcomer (30 - 50, new to adaptogens) Heard about adaptogens through a friend or article, overwhelmed by options, needs education and reassurance, price-sensitive on first purchase. Discovery channel: Google search ("what are adaptogens," "best supplements for stress").

• The Practitioner/Retailer (B2B) Naturopaths, wellness coaches, and independent retailers looking to stock or recommend quality adaptogenic products. Needs wholesale pricing, clinical backing, and brand credibility. Discovery channel: industry events, direct outreach, professional networks.

Each persona would be documented with specific behavioural triggers, marketing implications, preferred lead magnets, messaging angles, and objection handling frameworks.

Why this matters: Without validated personas, every subsequent decision — what lead magnet to create, what email to write, what scoring threshold to set — is based on pure assumption. Assumptions compound. By week 2, the entire system would be built on guesswork. We eliminated that here.

1.2 Lead Magnet Creation

Based on the validated personas, the content strategist moved on to develop three high-converting lead magnets, each mapped to a specific persona's most urgent problem. For instance:

Lead Magnet 1: "Stress Response Assessment — Find Your Perfect Adaptogen Protocol":
Target: The Overwhelmed Professional and The Curious Newcomer.
An interactive 12-question assessment that evaluated the user's stress patterns, current symptoms, lifestyle factors, and supplementation history, then delivered a personalised adaptogen protocol recommendation. This format converts at 30–45% in the health and wellness space — significantly higher than static PDFs (typically 8–15%) — because it provides immediate, personalised value. The assessment went on to include four result types (each mapped to a specific product recommendation), with dosage guidance, timing recommendations, and a "why this works for you" explanation citing relevant research. Each result page included a soft call-to-action to the recommended product with a first-purchase incentive.

Lead Magnet 2: "The Adaptogen Decoder — What the Science Actually Says":
Target: The Wellness Optimiser and The Overwhelmed Professional.
A 20-page guide that broke down the clinical evidence behind the eight most common adaptogens, explained mechanism of action in plain language, provided dosage ranges from published studies, and included a decision matrix for building a personalised stack. This positioned Rootwell as the brand that leads with science rather than marketing hype — a meaningful differentiator in a crowded supplement market.

Lead Magnet 3: "Wholesale Partner Kit":
Target: The Practitioner/Retailer.
A B2B-oriented package including product spec sheets, clinical study summaries, margin structures, marketing support materials, and a streamlined application process. This was a different lead generation path entirely — qualifying potential stockists and practitioner partners through a structured intake rather than leaving it to ad hoc email enquiries.

Landing page development for each lead magnet would follow a proven conversion framework: a headline formula combining specific outcome, timeframe, and obstacle removal; benefit bullets (5–7); trust indicators (testimonials, research citations, credential displays); a minimal-friction form (name and email only); and mobile-optimised design loading in under 3 seconds.

An honest caveat: Lead magnet conversion rates vary enormously based on traffic quality, brand trust, and offer relevance. The 30–45% range for assessments assumed warm traffic (social media followers, blog readers, retargeting audiences). Cold paid traffic converted lower — typically 15–25% for assessments and 5–12% for static content. We set expectations accordingly during Phase 1 and adjust projections based on Rootwell's actual traffic composition.

1.3 Multi-Channel Technical

The web developer built the technical infrastructure connecting lead capture to the rest of the system, as follows:

Landing page deployment: Three dedicated landing pages on a subdomain (://rootwell.com), built on custom HTML/CSS/JS templates for design flexibility. Each page was optimised for speed (caching, CDN configuration, image compression) with a target load time under 2.5 seconds — because every additional second of load time reduces conversion rates by approximately 7%.

Form integration: JotForms configured with anti-spam protection, connected to the email marketing platform (ActiveCampaign or Klaviyo, dependent on Rootwell's existing stack or preference) with field mapping for lead source, lead magnet downloaded, and persona type. Each form submission triggered an automated welcome sequence and CRM record creation.

Analytics configuration: Google Analytics 4 properly configured with enhanced e-commerce tracking, custom events for lead magnet downloads, email signups, and page engagement milestones. Conversion goals set with estimated lead values. Facebook Pixel installed with standard events (ViewContent, Lead, CompleteRegistration) to build retargeting audiences and enable lookalike targeting for future paid campaigns.

CRM integration: Whether Rootwell used HubSpot, Pipedrive, or a simpler tool, we configured automated contact creation, lead source attribution, persona tagging, and follow-up task generation for qualified leads.

Phase 2: Automation Implementation (Weeks 3-4)

2.1 Lead Scoring Automation

The marketing automation specialist designed and deployed a scoring model calibrated to Rootwell's specific sales cycles and customer behaviour.

Behavioural scoring (weighted by purchase intent signal strength):

Projected Outcomes

Demographic and firmographic scoring would add points for professional email domains (+5), geographic location in primary markets (+10), and indicators matching ideal customer profiles.

Threshold actions:
Score 0-20: Nurture sequence (educational content, no sales pressure).
Score 21-49: Engaged prospect (product-focused content, social proof, soft offers).
Score 50-74: Warm lead (targeted offers, case studies, urgency elements).
Score 75+: Hot lead (founder notification, personal outreach trigger, priority follow-up).

Implementation:
The scoring engine was run on one of the popular automation platforms, with webhooks from the website feeding behavioural data into the scoring model, which updated the CRM in real time. When a lead crosses the hot threshold, a notification fires to the founder with the contact's name, email, score, and most recent high-intent action.

Why we set these thresholds conservatively:
Many lead scoring implementations set the "hot lead" threshold too low, flooding the sales contact with lukewarm prospects and eroding trust in the system. At 75+ points, a contact has demonstrated multiple high-intent behaviours — they've engaged with content, visited product pages repeatedly, and shown sustained interest. This meant the founder's limited time was spent on conversations with the highest probability of conversion.

2.2 Automated Nurture Sequences

Based on the personas and scoring tiers, the content strategist and marketing automation specialist built persona-specific email sequences, like:
a. Sequence 1: Assessment Taker - Personalised Protocol Follow-Up (8 emails over 21 days)
This sequence was triggered when someone completed the Stress Response Assessment. The content is dynamically personalised based on their result type.
• Email 1 (Immediate): Your personalised results + detailed protocol explanation + "why this adaptogen works for your stress pattern".
• Email 2 (Day 2): "What to expect in your first week" — sets realistic expectations, addresses common concerns about adaptogens (onset timing, taste, interactions).
• Email 3 (Day 5): Deep dive into the science behind their recommended adaptogen — mechanism of action, key studies, dosage rationale.
• Email 4 (Day 8): Customer story matching their profile ("Meet Sarah, a project manager who struggled with the same stress pattern...").
• Email 5 (Day 11): "Building your complete protocol" — introduces complementary products, cross-sell opportunity.
• Email 6 (Day 14): FAQ addressing the top objections for their persona (price, "do supplements actually work," safety concerns).
• Email 7 (Day 18): Limited-time first-purchase offer with urgency (not artificial scarcity — genuine introductory pricing).
• Email 8 (Day 21): Final value-add, a downloadable "stress management toolkit" with non-product recommendations (sleep hygiene, breathing exercises, adaptogen timing optimisation), soft close.

b. Sequence 2: Guide Downloader — Education-to-Conversion (6 emails over 18 days)
For those who downloaded the Adaptogen Decoder guide — these leads are information-gatherers who need to build trust before purchasing.
• Email 1 (Immediate): Guide delivery + "how to use this guide" quick-start
• Email 2 (Day 3): "The 3 mistakes most people make with adaptogens" — positions Rootwell as the expert who prevents costly errors.
• Email 3 (Day 7): "Which adaptogen is right for your goal?" — interactive element driving back to the assessment.
• Email 4 (Day 10): Behind-the-scenes: Rootwell's sourcing and testing process — builds trust through transparency.
• Email 5 (Day 14): Social proof compilation — reviews, practitioner endorsements, before/after testimonials.
• Email 6 (Day 18): Personalised product recommendation based on engagement behaviour + first-purchase incentive.

c. Sequence 3: Wholesale Enquiry — B2B Qualification (4 emails over 10 days)
A tighter, more professional sequence for practitioner and retailer leads.
• Email 1 (Immediate): Partner kit delivery + founder introduction.
• Email 2 (Day 3): Margin structures, minimum order quantities, marketing support overview.
• Email 3 (Day 6): Case study of an existing retail partner's experience (anonymised or with permission).
• Email 4 (Day 10): Direct call-to-action to schedule a call with the founder.

What we decided not to do: We didn't build a 30-email drip campaign that takes months to play out. Supplement purchase decisions — particularly for a brand at Rootwell's price point — typically resolve within 2–4 weeks. Longer sequences see diminishing returns and increasing unsubscribe rates after email 8–10. We kept sequences tight, value-dense, and action-oriented.

2.3 Sales Handoff Automation

For a one-person operation like Rootwell, the "sales team" is the founder. The system needed to respect that constraint.

Automated handoff triggers:
• Lead score crosses 75+ → Slack notification with full context (name, email, score, lead magnet downloaded, pages visited, emails opened/clicked).
• Wholesale enquiry submitted → Immediate Slack notification + CRM task creation with 24-hour follow-up deadline.
• Reply to any nurture email → Notification with reply content for personal response.

CRM task automation: Follow-up tasks auto-created with priority levels, due dates, and pre-populated context so the founder doesn't waste time re-researching a lead before responding.

What this replaces: The founder currently checks email intermittently, responds to enquiries when they remember, and has no system for tracking who's been contacted and who's fallen through the cracks (p. 7). The automation doesn't replace human connection — it ensures human connection happens at the right moment with the right context.

Phase 3: Optimisation & Scaling (Ongoing)

After deployment, the system enters a monitoring and refinement cycle.
Weeks 5-6: Monitored all automations for errors, delivery issues, and unexpected behaviour (p. 8). Reviewed initial email performance metrics (open rates, click rates, unsubscribe rates) against benchmarks. Adjusted subject lines, send times, and content based on early data.

Weeks 7-8: First scoring model calibration. Compared scored leads against actual conversions — are leads scoring 75+ actually converting at higher rates than those scoring 40–50? If not, the weighting needs adjustment. This calibration was and is critical and is often skipped by agencies that deploy and disappear.

Ongoing monthly Dashboard review covered lead volume by source, conversion rates by funnel stage, email sequence performance, scoring accuracy, and revenue attribution. Quarterly strategic review with the founder to align lead generation priorities with business goals (for example, shifting emphasis to B2B wholesale if DTC growth plateaus).

Projected Outcomes: These projections were based on industry benchmarks for the health and wellness supplement category, published conversion rate data, and our experience with comparable implementations. They assumed Rootwell maintained its current traffic volume (18,000 monthly uniques) and did not significantly alter its product offering or pricing during the measurement period.We present ranges, not single figures, because honest projections acknowledge variability.

Lead Capture

Basis for projection: The 1.2% baseline reflected an unoptimised footer signup — the lowest-performing capture format. Adding targeted lead magnets with dedicated landing pages typically lifts conversion to 3–8% depending on traffic quality and offer relevance. The assessment format specifically converts at the higher end of this range for health and wellness audiences, based on published benchmarks from quiz platform providers and supplement industry case data.

What we did to make it lower: If Rootwell's traffic was predominantly low-intent (accidental clicks, bot traffic, irrelevant referral sources), conversion rates would've sat at the conservative end. A traffic quality audit during Phase 1 identified this.

Lead-to-Customer Conversion

Basis for projection: Email nurture sequences in the supplement space convert at 5–25% depending on list quality, content relevance, and offer structure. Our persona-specific approach with behavioural triggers sat at the higher end of published benchmarks. The 8–12% conservative figure accounted for the reality that many leads are early-stage researchers who may take months to purchase or may never purchase at all.

Revenue Impact

Basis for projection: At 18,000 monthly visitors converting at 4–7%, producing 720–1,260 leads, with 8–20% of those converting to customers at a $65 AOV, the monthly revenue addition ranges from approximately $3,700 to $16,400. The wide range reflects the compounding uncertainty across multiple conversion stages — each variable introduces variance, and the spread compounds at each step.

What this does not include: Revenue from paid traffic (which Rootwell was not running), B2B wholesale conversions (which have a different and longer sales cycle), or repeat purchases from newly acquired customers (which would compound over time but were difficult to project at 90 days).

Acknowledged Limitations of The Projections

1. These were projections, not guarantees. Every brand's audience, market position, and product-market fit is different. Projections were based on benchmarks and comparable implementations, but client results would be determined by factors including their product quality, pricing, competitive landscape, and customer service.
2. The 90-day window was short. Lead generation infrastructure compounded over time — the scoring model improved with data, nurture sequences got refined based on performance, and organic traffic grew as lead magnets generated backlinks and social shares. The 90-day projections captured the initial impact, not the mature-state performance.
3. These numbers assumed existing traffic. If Rootwell's monthly traffic dropped significantly (due to algorithm changes, seasonal fluctuation, or competitive pressure), lead volume dropped proportionally. The infrastructure captured a higher percentage of visitors, but it didn't create visitors.
4. Client cooperation is not optional. The projections assume the Client provides timely access, feedback, and approvals per the implementation timeline (p. 10). Every week of delay in Phase 1 pushes the entire timeline and the 90-day projection window forward.

The Investment

What the infrastructure fee covers: Everything described above — persona development, lead magnet creation (content, design, and landing pages), multi-channel technical setup (analytics, tracking, CRM integration), lead scoring automation, nurture sequence build, sales handoff configuration, and initial optimisation.

What it does not cover: Ongoing paid advertising spend, third-party software subscriptions (email platform, CRM, analytics tools), stock imagery or custom photography, and ongoing content creation beyond the initial sequences. These are the Client's responsibility, as detailed in our Terms and Conditions.

Performance bonus structure: The 25% bonus is calculated on revenue that exceeds the agreed baseline — meaning the Agency only earns more when the Client earns more. The baseline, attribution methodology, and measurement framework are agreed in writing during Phase 1 before any bonus calculations begin. This alignment of incentives is deliberate: we have no interest in inflating projections to close a deal, because our compensation depends on real, measurable results.

Why This Approach Works for Niche CPG

Generic lead generation playbooks — the kind built for SaaS companies or large retailers — fail niche CGP brands for predictable reasons:
• The audience is smaller and more specific. A supplement brand targeting stressed professionals is not working with millions of potential leads. The funnel needs to convert at higher rates because the top of the funnel is inherently narrower. That's why persona-specificity and high-value lead magnets matter more here than they would for a mass-market brand.

• Trust is non-negotiable People put these products in their bodies. The bar for credibility is higher than it is for a software purchase or a clothing brand The entire lead generation system — from lead magnet content to nurture emails — must establish scientific credibility and brand trust before asking for a transaction. This is why we lead with education (the Adaptogen Decoder, the Stress Response Assessment) rather than discounts .

• The founder is the bottleneck Most niche CGP brands are founder-led with small teams. The system must be automated enough to run without daily intervention, but intelligent enough to surface the right opportunities for the founder's limited time Lead scoring and automated handoffs are not luxuries for these brands — they're necessities.

• Repeat purchase is where the real value lives. The infrastructure we built captured first-time buyers, but the lead scoring and CRM data it generated became the foundation for retention and lifecycle marketing (which is a separate infrastructure solution). First purchase is step one; the system is designed to make step two inevitable.

Phase 1: Data Integration and Analytics (Weeks 1–4)

1.1 Customer Data Collection and Audit

The data analyst began with a comprehensive audit of every data source in Kava & Kin's ecosystem. Data source inventory:
• Shopify: 24 months of order history, customer profiles, subscription data, refund/return records, product catalogue.
• Mailchimp: subscriber list, email engagement history (opens, clicks, unsubscribes), campaign performance.
• Google Analytics: website behaviour data (assuming basic installation; likely requires reconfiguration for useful event tracking).
• Gmail: customer service interactions (unstructured, requires manual review of themes and sentiment patterns).
• Instagram: engagement data, DM history, comment sentiment.
• Subscription platform (e.g., Recharge or Bold): subscription lifecycle events, skip/pause/cancel data, payment failure records.

Data quality assessment evaluated:
• Completeness (what percentage of customer records have full contact information),
• Accuracy (are email addresses valid, are purchase records consistent between Shopify and the subscription platform),
• Consistency (date formats,
• Name formatting, duplicate entries), and
• Gaps (what critical data points are missing entirely).

Historical data extraction: We pulled a minimum of 24 months of purchase behaviour, email engagement, and support interactions. For a brand like Kava & Kin, we expected to find 30–40% data quality issues (incomplete records, duplicates, formatting inconsistencies) that needed resolution before any meaningful analysis can begin.

An honest caveat about data:
The quality of every output in this solution — segmentation, churn prediction, health scoring — was bound by the quality of the input data. If Kava & Kin's historical data was significantly incomplete or inaccurate, we may have needed to extend Phase 1 for additional data cleansing work, scoped and costed separately. We would identify this early and discuss it transparently rather than building models on a shaky foundation.

1.2 Behavioural Segmentation (RFM Analysis)

With clean, consolidated data, the data analyst implemented RFM (Recency, Frequency, Monetary) segmentation. RFM scoring for a functional beverage brand. RFM scoring for a functional beverage brand:
Recency:
• 5: Purchased within the last 30 days.
• 4: Purchased 31–60 days ago.
• 3: Purchased 61–90 days ago.
• 2: Purchased 91–180 days ago.
• 1: Purchased 180+ days ago.

For functional beverages, the expected repurchase cycle is 30–45 days for regular users. A customer who hasn't purchased in 60+ days is already showing risk signals. This is shorter than skincare (90–120 days) or supplements (60–90 days) because beverage consumption is daily.
Frequency (1–5 scale):
• 5: 10+ purchases.
• 4: 6–9 purchases.
• 3: 4–5 purchases.
• 2: 2–3 purchases.
• 1: 1 purchase (one-time buyer).

Monetary (1–5 scale):
• 5: Top 20% of lifetime spend.
• 4: 21–40%.
• 3: 41–60%.
• 2: 61–80%.
• 1: Bottom 20%.

Projected segments for Kava & Kin:
• Champions (RFM 555, 554, 544, 545): estimated 5–8% of base: Recent, frequent, high-value. These are loyal subscribers and multi-product buyers. Action: reward loyalty, request reviews and referrals, early access to new products, VIP treatment.
• Loyal Customers (RFM 543, 444, 435, 355): estimated 8–12%: Regular purchasers with solid value. Action: upsell tincture concentrates to RTD buyers, cross-sell across product ranges, subscription conversion offers.
• Potential Loyalists (RFM 512, 511, 422, 421): estimated 10–15%: Recent customers with growth potential. Action: educational content about building a daily ritual, product pairing suggestions, gentle subscription nudges.
• New Customers (RFM 511, 512): estimated 15–20% First-time buyers in the critical conversion window. Action: onboarding sequences, usage guidance, "what to expect" education, second-purchase incentive.
• At Risk (RFM 155, 254, 144, 214): estimated 12–18% Previously engaged customers showing decline. Action: retention campaigns, preference surveys, personalised re-engagement offers.
• Cannot Lose (RFM 155, 144, 215): estimated 3–5% High-value customers who have gone quiet. Action: direct founder outreach, significant re-engagement offers, personal check-in.
• Hibernating/Lost (RFM 111, 112, 211): estimated 25–35% Long-inactive, low-value. Action: low-cost win-back sequence, accept natural attrition for truly lost customers rather than wasting resources.

What we decided not to do: We didn't treat the "Hibernating/Lost" segment with the same intensity as "At Risk" or "Cannot Lose." A significant portion of one-time buyers were always going to be one-time buyers — they were curious, they tried it, it wasn't for them. Spending heavily to reactivate customers with no engagement signals is a poor use of resources. We focused investment on the segments where intervention had the highest probability of impact:
• New Customers (preventing drop-off),
• Potential Loyalists (accelerating repeat purchase), and
• At Risk (intervening before it's too late).

1.3 Predictive Churn Modelling

Our data analyst went on to build a churn prediction model calibrated to Kava & Kin's specific customer behaviour patterns.

Churn definition for functional beverages No purchase in 60 days (for non-subscribers) or subscription cancellation/3+ consecutive skips (for subscribers). This is tighter than the 90-day definition used for supplements because beverages are consumed daily — a 60-day gap represents a genuine behaviour change, not just a slow consumption cycle.

Behavioural risk indicators:
• High risk (immediate intervention): Email open rate decline exceeding 50% over 30 days; subscription cancellation attempt; product return or refund request; support ticket with negative sentiment; 2+ consecutive subscription skips.
• Medium risk (24–48 hour response): No website visit in 45 days; missed expected reorder date by 2+ weeks; reduced order value on consecutive purchases; email unsubscribe.
• Elevated risk (weekly monitoring): Gradual decline in email engagement scores; longer intervals between purchases; decreased product variety in orders (was buying 3 products, now buying 1).

Scoring algorithm: A weighted composite score (0–100) combining recency of purchase, email engagement trajectory, website visit frequency, support interaction sentiment, and subscription behaviour. Customers scoring above 60 were flagged for automated intervention. Customers scoring above 80 were flagged for manual outreach.

Model accuracy disclaimer: Predictive churn models are probability-based tools, not crystal balls. The model's accuracy depended entirely on the quality and volume of historical data. With Kava & Kin's 9,200-customer base and 24 months of history, we had enough data to build a functional model, but it needed 2–3 months of real-world calibration before its predictions became reliably actionable. During the initial period, we monitored predictions against actual outcomes and adjusted the weighting accordingly.

1.4 Customer Health Score Development

During Phase 2, the customer success manager and data analyst collaborated to build a composite Customer Health Score.

Health Score components for Kava & Kin:
Engagement Health:
• Email open rate (target: above 25%).
• Click-through rate (target: above 3%).
• Website session frequency (target: 2+ per month).
• Social media interaction (target: any engagement per month).

Purchase Health (35% of score):
• Purchase frequency vs. expected (target: on schedule ± 1 week).
• Order value trend (target: stable or increasing).
• Product range expansion (target: trying new products over time).
• Subscription adherence (target: below 5% skip rate).

Satisfaction Health (25% of score):
• Support ticket sentiment (target: neutral or positive).
• Review rating when given (target: 4+ stars).
• Referral behaviour (target: 1+ referral per year).
• Survey responses (target: NPS above 7/10).

What this enables: Instead of treating all customers identically, the Health Score created a real-time prioritisation system. The customer service coordinator — who managed a shared inbox reactively — gained a dashboard showing exactly which customers need attention, why, and what action to take. This transformed customer service from firefighting into proactive relationship management.

Phase 2: Personalisation and Automation (Weeks 5–8)

2.1 Automated Customer Journeys

The marketing automation specialist designed and deployed five core lifecycle journeys, each triggered by specific behavioural events and personalised by segment.

Journey 1: New Customer Onboarding (triggered by first purchase, 30-day sequence). This was the highest-leverage journey in the entire system. The gap between first and second purchase is where Kava & Kin lost the most customers.

• Day 0 (Immediate): Welcome email from the founders — personal, warm, not transactional. "Why we started Kava & Kin" origin story + what to expect from the product (onset timing, taste notes, how to incorporate into daily routine). Specific to the product purchased.
• Day 3: "Your first week with (Product Name)" — set realistic expectations for functional beverages (adaptogens need consistent use to show full effects), addressed common concerns (taste adaptation, optimal timing), and provided a simple daily ritual suggestion.
• Day 7: Ingredient deep dive — the science behind the key ingredient in their purchased product (e.g., "How kava root works: what the research shows"). Positioning the brand as the knowledgeable guide, not just a vendor.
• Day 14: Check-in + social proof — "How are you finding it?" with a link to a quick feedback form. Below the question, 2–3 customer testimonials from people with similar purchase patterns.
• Day 21 "Building your complete ritual" introduced complementary products based on what they purchased ("bought the focus elixir? Here's why pairing it with the evening blend creates a morning-to-night protocol"). Soft cross-sell, not a hard push.
• Day 28: Second purchase incentive — but not a generic discount. A personalised offer based on their engagement: high-engagement customers got early access to a new flavour; medium-engagement got a bundle offer; low-engagement got a straightforward discount + reminder of the product's value.

What we didn't do: We didn't send the same discount to every first-time buyer at day 28. Blanket discounting trains customers to wait for promotions and erodes margin. The tiered incentive structure rewarded engagement while still providing a conversion mechanism for less engaged customers.

Journey 2: Subscription Nurture (triggered by subscription activation, ongoing).

• Month 1: Subscription welcome + how to get the most from your subscription (customisation options, skip/swap flexibility, referral benefits).
• Month 2: Exclusive subscriber content — a recipe or ritual guide using their subscribed product.
• Month 3: Subscription milestone acknowledgement + preference check (still the right product? Right frequency?).
• Month 4: The critical churn-risk month (average subscription duration is 4.2 months). Proactive outreach: "We noticed you've been with us for 4 months — here's what long-term kava users typically experience" + upgrade or variety offer.
• Every 3 months thereafter Milestone celebrations, product recommendations based on evolving preferences, and genuine check-ins.

Journey 3: Churn Prevention (triggered by Health Score decline or churn risk score exceeding 60).

• Immediate: Subscription welcome + how to get the most from your subscription (customisation options, skip/swap flexibility, referral benefits).
• Day 1: Personalised check-in from the brand (not a sales email — a genuine "is everything okay?" message).
• Day 3: If no response, a value-reminder email: "Here's what (Product) has been doing for you" with educational content reinforcing the benefit.
• Day 7: Preference survey — "Has something changed? Help us help you" with options: wrong product, wrong frequency, budget concern, found alternative, other.
• Day 10 Based on survey response (or non-response), a tailored intervention: product swap suggestion, frequency adjustment, pause offer, or a direct discount as a last resort.
• Day 14 If still no engagement, graceful acceptance — a final message that respects their decision while leaving the door open: "We'll be here if you want to come back".

Journey 4: Win-Back (triggered by no purchase in 120 days for non-subscribers, or 60 days post-cancellation for subscribers). A 60-day sequence with declining frequency. The first touchpoint is the most important — if it doesn't land, subsequent emails have diminishing returns.

• Day 0: "We miss you" with a specific callback to their last purchase and what's changed since (new products, improved formulations, seasonal offerings).
• Day 14: Social proof from customers who came back — "Why (customer name) re-subscribed after a break".
• Day 30: Significant re-engagement offer (this is the one place where a meaningful discount is justified — the alternative is losing the customer permanently).
• Day 45: Final touchpoint — no sales pressure, just a "the door is always open" message.
• Post-60 Days Move to dormant list, reduce email frequency to quarterly at most.

Journey 5: VIP Programme (triggered by Customer Health Score above 85 or lifetime spend above $500). Ongoing programme for the brand's most valuable customers.

• Exclusive early access to new products and limited editions.
• Founder AMAs or virtual tasting sessions (quarterly).
• Referral programme with meaningful rewards (not a 10% coupon — a free box or exclusive product).
• Feedback loop: VIP customers get direct input into product development decisions.
• Surprise and delight: unexpected gifts on purchase anniversaries, handwritten notes on milestone orders.

2.2 Intervention Playbooks

Beyond automated journeys, the customer success manager would create playbooks for manual intervention scenarios.

Scenario: Customer submits complaint:
• Within 2 hours: Acknowledge with empathy and ownership. Pause all marketing automation. • Within 24 hours: Provide solution or timeline. Offer appropriate compensation (replacement, refund, credit). • 7 - 14 days later: Follow up to confirm resolution. Gradually reintroduce appropriate content.

Scenario: High-value subscriber attempts cancellation:
• Immediately: Offer alternatives before processing (pause, skip, swap product, reduce frequency). • If they proceed: Process without friction (making cancellation difficult creates resentment, not loyalty) (Hypothetic... p. 9). Send a genuine exit survey. • 30 days post-cancellation: Begin win-back Journey 4 with personalised messaging.

Phase 3: Launch & Optimization (Weeks 9–12)

Weeks 9-10: Systems integration and tracking. API connections between Shopify, email platform, automation platform (Make.com or n8n), and CRM (HubSpot or equivalent). GA4 enhanced e-commerce configuration with custom events for lifecycle stage transitions, subscription events, and content engagement. Webhook-based real-time data synchronisation.

Weeks 11-12: Soft launch, testing, and training. All automations activated with close monitoring. Test customers run through each journey to verify trigger conditions, content delivery, and handoff logic. Team training on the Customer Health Score dashboard, intervention playbooks, and escalation procedures. Documentation delivered. Transition to ongoing Management Services.

Projected Outcomes These projections assumed Kava & Kin maintained its current customer acquisition rate, did not significantly alter its product line or pricing, and cooperated with the implementation timeline. We presented ranges because the outcomes depended on data quality, customer responsiveness, and execution consistency.

Customer Retention

Basis for projection: The 23% repeat rate reflected zero post-purchase nurture. Industry benchmarks for consumable DTC brands with automated onboarding sequences showed repeat rates of 35-50%, depending on product quality and sequence effectiveness. The conservative range accounted for the possibility that a significant portion of one-time buyers were truly not the right fit for the product. Subscription duration improvements were based on the compounding effect of engagement scoring + proactive intervention + preference management — each individually contributed 1-2 months of retention extension.

What we did to make it lower: If Kava & Kin's product had a fundamental taste or efficacy barrier that no amount of education can overcome, the retention improvements would be modest. The data from Phase 1 would've revealed this — if return rates were high or NPS is below 40, the problem is product, not marketing infrastructure .

Customer Lifetime Value

Basis for projection: CLV improvement comes from three drivers: (1) more repeat purchases (retention),
(2) higher average order values (cross-sell and upsell through personalised journeys), and
(3) longer subscription durations. Each driver is independently supported by the system — the projections reflect their combined, compounding impact.
The wide range on optimistic subscriber CLV ($260-$340) reflects the exponential nature of retention improvement: each additional month of subscription adds linearly to CLV, but the probability of reaching that month increases non-linearly with effective engagement.

Revenue Impact

Basis for projection: Churn-prevented revenue is calculated as: (number of subscribers who would have cancelled under current conditions) × (additional months retained) × (average subscription value). At 680 subscribers with ~18% monthly churn, approximately 122 subscribers cancelled per month. Reduced churn to 10-13% prevented 34-55 cancellations per month. At $29.50/month per subscriber and an average of 3-5 additional months retained, the annualised value ranges from $85K to $260K. This was revenue that would have been lost — it didn't require new customer acquisition.

Acknowledged Limitations of The Projections

1. The 186% CLV increase and 287% retention improvement cited in our proposal are optimistic ceiling projections: The conservative floor was 47-76% CLV improvement and 48-65% retention improvement. Both are meaningful — but the spread between conservative and optimistic was wide because lifecycle optimisation outcomes depended heavily on data quality, product-market fit, and customer responsiveness.
2. Churn prediction models need calibration time: The churn model was not fully accurate at launch. It required 2-3 months of real-world prediction-versus-outcome comparison to calibrate properly. During that window, some at-risk customers were going to be missed, and some healthy customers were going to be incorrectly flagged. This is normal and expected — not a failure of the model.
3. Win-back success rates decline over time: A customer who churned 60 days ago had a meaningfully higher reactivation probability than one who churned 180 days ago. Our projections assumed the win-back infrastructure was deployed promptly after Phase 3, capturing recently churned customers while they were still reachable. Delay eroded the addressable win-back audience.
4. Subscription improvements require product-side flexibility: If the subscription platform didn't support pause, swap, frequency adjustment, or flexible SKU selection, the retention strategies were limited to communication-only interventions. We identified platform constraints during Phase 1 and recommended subscription platform migration if necessary (scoped separately).
5. This is the most expensive solution for a reason. The $20K infrastructure fee and $7K/month management fee reflect the depth of data work, model building, and multi-journey automation involved. For a brand with $1.8M in revenue, this is a significant investment. We believe the projected ROI justifies it — but we also believe in being transparent about the commitment required.

The Investment

What the infrastructure fee covers: Everything described above — customer data audit and consolidation, RFM segmentation, churn prediction model development, Customer Health Score design, five automated lifecycle journeys, intervention playbooks, systems integration, analytics dashboard deployment, and team training.

What the Management Services fee covers: Churn model recalibration and threshold adjustment, Customer Health Score monitoring and intervention management, journey performance analysis and optimisation, weekly reporting with segment-level insights, monthly strategic review, and automation maintenance.

What it does not cover: Third-party software subscriptions (CRM, email platform, subscription management platform, analytics tools), customer service staffing, product changes or reformulations, data migration from legacy systems not accessible via API, and paid advertising. These are the Client's responsibility.

Why the 6-month management minimum: Unlike content marketing (3-month minimum) or lead generation (no mandatory management period), lifecycle optimisation requires extended model calibration, seasonal pattern recognition, and iterative journey refinement. Meaningful statistical significance on retention improvements requires at least 4-6 months of data. Cutting management services at month 3 means cutting before the system has proven itself.

Performance bonus structure: The 35% bonus is calculated on revenue above the agreed baseline - specifically on improvements in customer retention, repeat purchase rates, CLV growth, and reactivated churned customers. Revenue from new customer acquisition is excluded unless directly attributable to referral programmes built within the Infrastructure. The baseline methodology and attribution framework are agreed in writing during Phase 1 before any bonus calculations begin.

Why This Approach Works for Niche CPG:

• Niche CPG brands can't rely on constant acquisition. The total addressable market for adaptogenic functional beverages is smaller than mass-market soft drinks. When you've reached 30-40% of your addressable audience through acquisition channels, growth has to come from making each customer more valuable - buying more frequently, for longer, across more of your product range.
• Consumable products have natural repurchase cycles. Unlike a one-time purchase (furniture, electronics, etc.), beverages are consumed daily and need regular replenishment. This creates a predictable rhythm that automation can anticipate, support, and optimise. A customer who should reorder every 30 days but hasn't by day 35 is giving a clear signal; the system detects it and responds before day 60, when the customer has likely switched to a competitor or simply stopped using the product.
• Subscription is the business model, but retention is the business. Kava & Kin's subscription cohort generates predictable monthly revenue - but only as long as subscribers stay. Improving subscription duration from 4.2 to 7.5 months doesn't sound dramatic, but it increases lifetime subscription revenue by 79% without acquiring a single new customer. At scale, this is transformative.
• Data is the underused asset. Most niche CPG brands are sitting on 2+ years of transactional data, email engagement data, and customer service interactions that have never been consolidated, analysed, or acted upon. The lifecycle platform doesn't create new data - it unlocks the value of data that already exists.