Case Study · 09 / 10 Distribution · Cross-sell Save as PDF ↓
Customer intelligence · 8-week build

The quiet goldmine.

How a Pune ingredients distributor pulled ₹1.9 Cr/month out of accounts they were already serving.

ClientTriveni Foodworks LocationPune, India PublishedMay 2026
Industry
Specialty ingredients · F&B
Scale
800 mfrs · 14 inside sales reps
Team
140 employees · ~₹100 Cr ARR
Engagement
8 weeks · ERP + CRM layer
Outcomes · 2 quarters post-launch
+₹1.9Cr
Incremental monthly revenue from existing accounts
6.3×
Dormant accounts reactivated per quarter (14 → 89)
−71%
Account research time per rep per week
§ 01 · The company

9 years of order data. And no way to use it.

The company

Triveni Foodworks is a Pune-based specialty ingredients distributor supplying flavors, emulsifiers, and preservatives to roughly 800 packaged food manufacturers across western and southern India. 140 employees, ~₹100 Cr annual revenue, 14 inside sales reps, and an ERP groaning under 9 years of order history.

Aniket Kulkarni, Head of Inside Sales, watched ~60% of monthly revenue come from the same top 160 accounts every month while the remaining 640 quietly ordered the same two SKUs and nothing else. What kept him up: not knowing which dormant account was about to churn — and not having time to find out.

§ 02 · The problem

Cross-sell was tribal knowledge.

Sales motion

Triveni had 9 years of order data and no way to use it. Reps spent ~7 hours per week digging through the ERP to prep account calls — mostly for the big 20% they already knew well. The bottom 60% of accounts got one check-in per quarter, sometimes less. Cross-sell was tribal knowledge: senior reps remembered which customer bought what, juniors did not.

Breaking pointOct 2025

A ₹38L/year biscuit manufacturer in Hyderabad churned to a competitor for an emulsifier Triveni already stocked. Nobody had told them. The post-mortem surfaced 47 other accounts in the same pattern: stable orders on 2–3 SKUs, zero outreach in 6+ months, complementary products they almost certainly needed.

₹4–5 Cr annual leak · one email away
§ 03 · The solution

A Monday-morning customer intelligence layer.

Architecture

Vynara built a customer intelligence layer on top of Triveni's existing ERP and CRM. It pulls transaction history every Monday morning, clusters accounts by industry and buying pattern, and flags two things per rep: dormancy risks (accounts deviating from their normal order rhythm) and cross-sell openings (SKUs that look-alike accounts buy but this account does not).

A second agent drafts the outreach — a 4-line WhatsApp or email message in Triveni's voice, referencing what the customer last bought and what they're likely missing. Reps approve, edit, or reject in a single queue. Detection and drafting became automated. The call, the relationship, and the close stayed with the rep.

Reps still own
the call · the relationship · the close
§ 04 · How it works

From "skim and remember" to "edit and send."

Before · Monday morning 7.2 hr / wk
  1. Rep opens ERP Monday morning
  2. Manually pulls report for ~50 assigned accounts
  3. Skims order history, tries to remember context
  4. Picks 5–6 accounts that "feel" worth calling
  5. Drafts each email from scratch
  6. Same top 20% accounts get called every week
After · Monday morning 2.1 hr / wk
  1. Rep opens Vynara dashboard Monday 9 AM
  2. Sees 12 prioritized accounts · scored by ₹ opportunity + churn risk
  3. Each shows: last order, suggested SKU, lookalike justification, draft message
  4. Rep edits draftavg 40 sec · sends
  5. Replies route back into the system to refine scoring
§ 05 · The numbers

Six metrics. Two quarters.

Monthly revenue from existing accounts+22.6%
Before₹8.4 Cr
After₹10.3 Cr
Research time / rep / week−71%
Before7.2 hrs
After2.1 hrs
Dormant accounts reactivated / qtr+6.3×
Before14
After89
Avg SKUs per active account+47%
Before3.2
After4.7
Outreach msgs / rep / week+3.5×
Before13
After46
Reply rate on outreach+127%
Before4.1%
After9.3%
§ 06 · The timeline

8 weeks. Then weekly tuning in-house.

Phase
W1
W2
W3
W4
W5
W6
W7
W8
Week 1ERP + CRM audit · schema mapping · access setup
AUDIT
Week 29-yr ingest · 640K rows · cleanup
INGEST
Week 3Clustering · lookalike model · dormancy logic
CLUSTER
Week 4Cross-sell engine · opportunity scoring
SCORE
Week 5Outreach drafting · brand voice tuning
DRAFT
Week 6Pilot with 3 reps · north + south territories
PILOT
Week 7Feedback · dashboard polish · full rollout
ROLLOUT
Week 8Handover · training · in-house tuning cadence
HANDOFF
§ 07 · What they said

The data was always there.

"

We always knew the data was there. We just didn't have 14 analysts. Now every rep walks in Monday with a list that used to take them half a day to build, and they spend that time actually selling.

AK
Aniket Kulkarni
Head of Inside Sales · Triveni
"

I closed a ₹6 lakh order in week three from an account I had not called in 8 months. The system told me they should be buying our acidulant. They were buying it elsewhere. Switched in one call.

RS
Rohit Sharma
Senior Sales Executive · Triveni
§ 08 · The takeaway

The goldmine is already paid for.

If your company has 3+ years of transaction history and a sales team that runs on memory and gut, the goldmine is already paid for.

The work is making it visible — account by account, every Monday morning. Vynara builds this in 6–8 weeks on top of whatever ERP or CRM you already use.

One-liner Your existing customers are buying from someone else what they could be buying from you. We can show you who, by next Monday.
Share your account list →