Building India’s Model of Agent Banking

In a small town called Canning, about two hours away from Kolkata, India, I meet DK Sardar, a bank agent, and follow him to his village called Baheesh Shona on a dirt road that makes a journey of ten kilometers take thirty minutes. Sardar’s house is located on the arterial road within the village. Twenty-eight years old and a recent college graduate, he leads us inside his house where his father has just finished rolling hundreds of fresh bidis (small cigarettes) for sale. Along with small-scale farming, his family runs a bidi-making business. Sardar lays out a biometric device, a mobile phone, and his agent certification from the State Bank of India and explains that even though he enrolled a hundred people six months ago when he was first certified as an agent, the local bank branch has not yet made any of the accounts active.

As a result the initial queues of people lining up outside his house to open accounts have dried up. He has invested Rs 50,000 (about $1,000) to buy materials to build an office, and for the deposit he needed to pay for the equipment. Repeated visits to the bank branch by him and the affiliated company that provided him the equipment has not sorted this problem out yet.

CGAP, in collaboration with the College of Agricultural Banking, just completed a national survey, which captured the big picture on agents across the country. In India, the term customer service point (CSP) is used to refer to individuals who act as agents on behalf of banks. Business Correspondents (BCs) are companies that source and manage one or more CSPs on behalf of banks. The use of CSPs is still new in India although there are about 80,000 such points nationally. The recent expansion means most CSPs have been operating for less than 24 months. Eighty five percent are in rural areas and many in villages that are targeted under the national financial inclusion plan.

The survey focused on the value of this service to customers, and motivations for CSPs to work and continue working. A summary of survey results is being released through the attached PowerPoint presentation, developed by Greg Chen and Aimthy Thoumoung. Deeper analysis on the results will be made available later in August.

The survey results show that approximately one-quarter of CSPs are either unavailable or unable to transact – perhaps for reasons similar to ones Sardar encountered. About one-third of clients are limited to using one CSP. India is a bit different than other countries in that 30% of CSPs are roaming or “moving points” which has advantages and disadvantages which need to be understood better. The survey says that the time taken by CSPs to complete account opening paper work is reasonable, but account activation in many cases by BCs and Banks, like in the case of DK Sardar described above, is too slow. Most CSPs focus on single (often payments) product and do relatively little cross-selling.

Qualitatively, CSPs cite that the job provides an elevated status to them within their community, which is a significant boost to their motivation. However, data from the survey show that CSP income is low in most cases and may not be sufficient to maintain motivation. As Sardar pointed out, the primary reason for him to work as a CSP is getting an income. The second is providing a service to his community.

The challenges illustrated by Sardar’s story and the CSP survey are indeed the challenges of building an active, high quality network of CSPs. Quality needs to become a focus in government targets, contracts and performance metrics. Support to CSPs from BC Companies, and Banks, needs to be improved and made more consistent. And training, pay, regular visits and more reliable technology are needed.

As I leave Baheesh Shona village, Sardar points to the bricks he has bought to build his office, “I was hoping to earn Rs 10,000 ($200) per month but nothing has come through yet. I have no choice but to wait, since I have committed a lot of my family’s resources into this investment.”

Sardar’s story provides a window into India’s financial inclusion picture. There is strong will at the policy level, plenty of good ideas, robust demand for services, and motivation at the ground level to get the job done. But in the vast maze of India’s financial architecture, it is a challenge to implement the big ideas while keeping track of detail and quality. Agent networks are expanding rapidly in India but Sardar’s story and the CSP survey show that it’s more important to focus on building quality, rather than growth in the next phase of developing branchless banking in India.

-Shweta S Banerjee-


CSP Survey Results can be accessed here.


You can follow CGAP Technology Blog’s special series on India’s Branchless Banking scene, where key issues, and innovation by Eko,and Beam have been discussed.


13 August 2012 Submitted by Anonymous (not verified)

This analysis largely reflects the results from MicroSave’s extensive work on 20 agent networks across India using its Agent Assessment tool.

For MicroSave’s Policy Brief on the State of Agent Networks:…
Also watch the video where the business correspondent agents talk about the problems of business viability and insufficient business case:…

While the rural agents rarely see significant business, the urban remittance market can be profitable and becoming increasingly competitive. See:…

Also, watch the videos where the business correspondent agents talk about the various service providers offering remittances and what is important for the agents as well as the customers: and…

13 August 2012 Submitted by Anonymous (not verified)

Hi Sweta!

At the very outset, lets accept the fact, that, this agent model, has addressed the access to basic financial banking services to doorstep, in locations, where, depositing/withdrawing Rs 100 incurred a cost of at least direct Rs 35-40.
Bus fare – Rs 10-20 (up n down)
Lunch/meal – Rs 10
Indirect Opportunity cost lost – Rs 100-150 (full day of wage is lost, and goes for travelling to nearby Bank branch)

So, the model, has at least given one unparalleled fact “Convenient Banking” to the clients, which was just unthinkable ten years back.

This agent model, needs to incorporate two basic facts :
1. Scalability
2. Business sense (to be crude – profit)
We have seen, that a long queue in wee hours of business, at an agent location, sometimes, takes a toll on the agents main business. Why should have an agent break his head for dealing with clients for 10-15 min, and explaining him account opening formality and finally get just meagre .5-.8 % of transaction value.

Secondly, many Banks stipulate that the agent location can only cater to one Bank, kind of “clause of non compete”, which reduces any further chances of diversified cash flows.

Banks need to give these agents diversified product basket, an attractive commission structure, which makes business sense. Lets accept the fact, that , BC model was never started to achieve “scale”, but since the goals have changed mid way, with roaring success, accordingly, Government/Banks should make these agents, a proper shareholder of benefits. Lets not forget, that, agent save a hell lot of fixed cost to Banks (physical space).

Agents are the most trusted lieutenant in their areas among locals, so need of hour, is to incorporate some elements of “true business proposition” in this model, other wise, this will again remain a “half baked cake” in India Financial Inclusion oven.


Add new comment