How Rupee112 Reduced Drop-Offs by 3% using Weya AI
Feb 4, 2025
Rupee112, an NBFC based in Gurugram, was facing a common yet painful problem: too many prospective borrowers were dropping out during the loan sanction process. Key stages like the collection of documents or eKYC weren’t being completed, and many applications simply never got to the disbursement stage. The result was lost opportunities, wasted marketing spend, and a less efficient operations pipeline. Weya AI came in to help transform that process using automated workflows, timed nudges, and multi-channel conversational outreach to guide customers through every sticking point. The outcome: a measurable improvement in loan sanction conversion, lower cost per conversion, and a smoother customer experience.
The Problem: Drop-Offs in the Sanction Stage
Here’s how Rupee112’s customer journey was structured (simplified):
Stage | What the customer does | What often fails/causes drop-off |
1. Basic info capture (name, email, phone, address) | Customer fills basic form → profile created if they meet basic eligibility | Some customers never complete the basic form; others are rejected here. |
2. Document & verification stage (Aadhaar, PAN, bank statements, eKYC via Digilocker, etc.) | Once the profile is created, the customer must submit the required documents + complete eKYC | Many customers stall here — confusion, lack of clarity, forgetfulness, busy schedules, or unsure how to use Digilocker. |
3. Follow-up & conversion | The customer moves through to the sanction and loan disbursal | Without regular touchpoints, many drop off at stage 2; manual calling was attempted, but wasn’t scalable or reliable. |
Rupee112 needed to reduce friction at stage 2: improve follow-up, provide guidance, and increase the rate at which document submission and verification were completed.
Weya’s Approach: Intelligent Nudges, Multi-Channel Outreach & Automation
Here’s how we designed the solution:
Workflow Triggers for Drop-Off Detection
Time-based triggers in Weya workflows: e.g., if a customer doesn’t submit a document within X hours/days after profile creation, trigger a “nudge.”
These nudges aren’t generic — content tailored to their current missing requirement: “Hi, you’re almost there — just need your Aadhaar / PAN to move ahead.”
Voice AI Follow-Up
Automated calls (voice agents) to customers to explain what’s needed: which documents, how to eKYC via Digilocker,and why it matters.
Voice helps especially when people are busy or not responding via texts; allows for more natural conversation and clarifications.
WhatsApp AI / Messaging Channel
For those who don’t pick up or respond to voice calls, Weya’s system shifts to WhatsApp messages.
Friendly reminders: “Hi, are you available to complete your application now? We can chat here.”
These messages act as follow-ups/nudges in a less formal, lower-barrier channel.
Guidance Built Into Each Touchpoint
Every outreach (call or message) includes precise instructions: how to upload documents, how to use Digilocker, etc.
Proactive help: “If you have trouble with X, reply here and we’ll assist.”
Lower Cost, Higher Frequency
Automation reduces dependence on manual calling teams.
Because outreach is triggered dynamically, frequency of touchpoints increases without proportional increase in manual effort.
The Results: What Rupee112 Saw
These are the key outcomes after deploying Weya across their sanction process:
Loan sanction rate improved by ~ 3% over their baseline.
Lower cost per successful sanction (because many follow-ups are automated rather than manual).
Better throughput at document / eKYC stage: fewer applications being abandoned, more complete profiles.
Improved customer satisfaction / fewer complaints, since guidance reduced confusion.
Why It Worked: Key Success Factors
Here are what I believe were critical in making this transformation succeed:
Timely & Relevant Nudges
Customers dropped off because they didn’t know what to do or when. Delivering reminders exactly when they’re needed — not too early, not too late — makes a difference.Channel Flexibility
Voice calls capture engagement in certain moments. Messages / WhatsApp pick up where calls don’t. Multiple channels ensure fallback.Guidance, Not Just Reminders
Many users were stuck not due to lack of trying, but due to not knowing how to complete the tasks (e.g. how to upload or use eKYC tools). Clear, simple instructions helped.Automation Scaling Without Losing Personal Touch
Although the system is automated, the content of outreach felt personal (“you”, “we’re here”, helpful). That maintains trust and reduces friction.
Lessons Learned & Best Practice Tips
From working with Rupee112, these lessons can be applied broadly by NBFCs, fintechs, or any organization with multi-step customer journeys:
Map where drop-offs happen (quantify them) → decide triggers around those points.
Use voice + messengers + app notifications in tandem. Don’t rely on a single channel.
Ensure every communication enables the customer: include “how to” help, not just “you need to do X”.
Automate repetitive follow-ups but monitor responses so you can tweak messaging or timing.
Track ROI per-channel to see which mode of outreach (voice vs message) yields better conversion per cost, then optimize.
Conclusion
For Rupee112, leveraging Weya AI’s workflows, voice agents, and WhatsApp nudges turned a leaky sanction funnel into a much more efficient, guided journey. A 3% increase in sanction rate might sound modest— but in high-volume lending, that implies significantly more approved loans, more revenue, and less wastage of acquisition cost.
At Weya, we believe every customer who starts a loan application deserves to finish it — with minimal friction. If you’re an NBFC or fintech seeing high drop-off in your sanction process, reach out. Let’s build a customized workflow that turns more prospects into customers, more quickly, more humanly.
