A regular dose of D2C-centric resources & tools for Growing Brands, Startups & Entrepreneurs.

Every morning across India, the ground outside homes comes alive with patterns — drawn by hand - Rangoli, Kolam, Alpana, Muggu 🤷‍♂️ goes by many names. 

Patterns play a big role in India, understanding them? Even more so.

Over the past few years, we’ve worked with 900+ Indian D2C brands. And in that journey, what stands out isn’t just growth, it’s repetition.

Like the rangoli patterns drawn, the data tells a story. Let’s trace it.

Past ⏩ PATTERN ⏩ Present

Check out the Top 10 Patterns our D2Cs learn and benefit from every day…

1. Carrier Optimisation Using Peer Category Data

Let’s start simple:

Two D2C skincare brands ship to the same pincode, Brand A uses Carrier X, Brand B uses Carrier Y. One clocks 89% SLA adherence, the other 61%. Guess which brand also has a 23% higher RTO?

Now, imagine being able to predict and benchmark 😀 

This isn't magic — it’s pincode-level carrier prediction based on:

• Category-adjusted packaging sizes

• Payment method preferences by region

• Carrier handover latency by warehouse

One brand in Jaipur switched 128 pincodes to alternate carriers, optimised based on cross-brand data, and saw:

• SLA adherence jump by 16.2%

• RTO drop by 28.5%

2. Prepaid Nudges Engine via TOFU Flow Analysis

TOFU flows (Top-of-Funnel) are full of friction. Especially when trying to convert COD to prepaid.

But what if we knew:

• What exact time slots converted best by user region?

• What UPI nudges work in skincare vs electronics?

• What cashback ceiling nudged highest COD-to-Prepaid switch?

Using data from 900+ D2Cs, we built:

• Intent-predicted nudges at checkout page (abandoned last week? Offer ₹15 UPI cashback)

• Dynamic bank offer recommendations (based on user history)

Results:

• Prepaid share moved from 28% → 46%

• CVR (Conversion Rate) up by 19%

• RTO fell from 34% → 21%

3. NDR Pattern Matching with Behavioural Tags

Most brands treat NDR as a post-facto report. We treat it like a predictive category.

Based on over 58M+ NDR events (from 900+ D2C), we classify NDRs into:

• High Intent - Payment Issue (e.g., "No change available")

• Medium Intent - Absenteeism patterns (e.g., cluster of missed deliveries post 6PM)

• Fake NDR - Courier-triggered (rare but consistent)

Example: A jewellery brand saw 17% NDR in Bengaluru for COD orders. Deep dive showed:

• 82% of NDRs came from 4 courier partners.

• Same partners had <5% NDR with other brands.

Using cross-brand courier behaviour data, we triggered rerouting logic based on courier ID — and reduced false-NDRs by 66% in 14 days

NDR Classification Logic (Simplified)

4. COD Success Score (CSS) Based on Behavioural + Logistic Signals

We’ve built a scoring model — CSS (COD Success Score) — on 67.9M+ COD orders:

• Click–cart–order lag time: longer = more cancellations

• Device type: Android budget devices = 1.6x more likely to fail

• Pincode + previous NDR + average return window data overlay

CSS Output: Score range: 0 to 1

• 0.75 = safe to accept

• 0.5–0.75 = add verification step

• <0.5 = flag or convert to prepaid only

Impact:

• COD rejection down by 31.7% for filtered orders

• RTO loss reduced: ₹14.2L/month for mid-size brand in North India

5. Geo-Cohort LTV Clustering for Targeted CAC Allocation

Most brands deploy CAC like it’s a monolith. But acquisition in Surat ≠ acquisition in Pune.

Using cohort + geo data, we derive:

• LTV delta by city-pincode-category

• Repeat purchase windows by cohort size

• Ad-to-first-retention-event gap

Result: A D2C supplement brand running Pan-India FB ads saw:

• CAC: ₹224

• Blended LTV after 60 days: ₹488

After geo-cohort LTV insights:

• CAC (Optimised Tier 1 + UGC ad variants): ₹198

• LTV (Optimised cohort clusters): ₹712

CAC vs LTV: Before vs After

6. Cross-Brand Fraud Ring Detection Using Entity Graphs

Across 900+ brands, fraud rings tend to recycle tactics:

• Repeat pincode clusters

• Order timing patterns (1–2 AM spikes)

• Behavioural signals: COD-only, dummy names, alternate phone format

Model Input:

• Shared addresses + phone patterns

• NDR repeat logic

• Checkout funnel breakage points

Case: We flagged a group of 174 phone numbers triggering 3.2K fake COD orders monthly across 9 brands.

Action:

• Joint blacklist = 93% block success

• Shared model retraining = - 27% false positives

7. Micro Time-Slot Based Fulfilment Routing

One of the most ignored levers in D2C ops: Time-of-day based fulfilment routing.

Why does it matter? Because:

• SLA breach likelihood increases by 23% post-4PM pickup in Metro cities

• Courier performance shifts across weekday–weekend transitions

• Average OTP delay is 12.3 minutes longer in post-lunch slots in Tier-2 hubs

So we optimised pickup-slot allocation for brands based on:

• Courier efficiency per hour by warehouse region

• Last-mile congestion probability

• First-attempt delivery success by pickup time

8. Tier-2 Personalisation Models Using Regional Behavioural Vectors

Brands often stereotype Tier-2 buyers as ‘value-seekers’. Data says otherwise. From 17.1M orders across 103 cities:

• Higher CVR on mobile-optimised checkout in Hindi (Lucknow, Indore)

• Strong preference for bank-based UPI (vs. wallet)

• Longer browsing time → higher intent in skincare, not electronics

Case: A women’s wellness brand switched:

• Checkout copy to Hindi + Marathi

• Added PayTM + GPay fallback

Lift:

• Drop-off rate reduced by 22%

• CVR uplift: 2.3x in Tier-2/3 regions

Checkout Optimisation for Tier-2

9. Post-Order Journey Segmentation via Behavioural Traces

After delivery, there’s silence — but not for us. Using data from 11.2M post-purchase sessions:

• 68% of post-order page views happen within 24h of delivery

• “Where’s my order” (WISMO) drop-off correlates 0.63 with NPS improvements

• Repeat visit post-delivery = +21% higher LTV

Segmentation Flow Examples:

• “Trackers” → revisit tracking page 2–4× → nudge for review

• “Dormants” → zero activity → drop personalised WhatsApp reminder (not push)

• “Browsers” → open upsell messages within 48h → trigger loyalty onboarding

• “Previously Loyal” → has not engaged in a while → ask about previous order (optimise)

10. Refund Analysis: Dynamic Policy Based on Category × Location × Intent

Flat refund policies? That’s like using the same sunscreen in Delhi and Cherrapunji.

• Inner-city COD fashion = highest refund rates (Delhi South: 21.9%)

• Delay-induced refunds spike after 72h breach, especially in gifting

• Nutraceuticals in Tier-2 prepaid = lowest refund rates (<2%)

Smart Refund Rule Example:

• Category: Fashion

• Location: Tier-1 (Delhi, Mumbai)

• RTO probability: >18%

• → Enforce stricter image-based evidence + shorter refund window + exchange only options for select risky customers

Impact:

• Refund abuse down 28.6%

• CSAT steady (due to well-placed messaging)

That’s the end of our talk on “Your D2C & Data from 900+ D2Cs.”...

☕ See you on the next coffee date!

How did you like our Newsletter? What topic would you like us to cover next?

Reply and let us know! 😊