Project and system description – software system for small farmer microfinance (Kenya) with capability for global roll-out

http://www.cd3wd.com/sfmss/ - Small Farmer Microfinance Software System

http://www.cd3wd.com/sfmss/sfmss.zip - a zip file of the word document – and now including the source code for the postgressql system…

http://whiteafrican.com/?p=414

The system handles automatic and semi-automatic communication and payments between farmers’ groups (FG), the project management organization (PMO), commercial and semi-commercial stockists of farmers’ inputs, and the produce purchasing company (PPC).

The system relies very largely on structured gsm mobile phone sms’s (text messages) which are sent by FG’s and by stockists and the PPC’s crop collection truck drivers, and these are automatically processed by the software system, to rules and parameters set by the PMO management team. They result in payments being made, distribution and/or collection schedules being organized, outgoing emails and sms’s being actioned etc.. In the few cases where incoming sms’s are not correctly structured then they are flagged and raised for manual intervention and re-submission by PMO staff (either with or without human mobile phone communication to FG or other actor).

Using this system it is anticipated that a staff of 2 can handle 1000 groups, 20,000 farmers. The resultant high level of organization and low costs will produce a good overall system for farmers, FG representatives, stockists, PPC’s etc..

The system uses an open source Postgresql database with heavy use of functions – i.e. a database-intensive, application-light programming approach. The client robot application and the various management interfaces are written in java (and the interfaces are web-based).

The system is very heavily parameterized, so as to give great power to the PMO management team, and also to make the system transportable to a number of countries and languages.

There are daily, weekly etc reports auto-produced and auto-emailed which enable the PMO management team and the buyer to spot and handle trouble before it occurs.

The system uses an international sms gateway service, currently located in India, and uses protocols (email2sms, sms2email, and http or https sms send) which could allow relatively easy switching to an alternative supplier, if technical, service-level or cost reasons so dictated.

The system operates in conjunction with electronic banking. At present the interfaces between the sfmss and the ebanking system are only semi-automated, we are working on raising that level of automation. Sfmss automatically reads and imports bank statement text files produced by the accountant on his or her ebanking web interface; sfmss also produces payment lists from which fields can be copied and pasted into the ebanking web interface payment screen. After such a manual payments session, the bank statement file can be created and re-imported to sfmss, and sfmss will feedback to the accountant whether all payments were correctly effected or not.. By 2007/06, the payments should be done from a sfmss-created file, which is submitted to the ebanking web interface.

To give a very cursory idea of that is going on in the system, the following steps occur:

- FG’s sign up and pay membership fee

- FG’s inform PMO total acreage

- PMO informs FG’s quantities of fertilizer and seed required and the resulting loan. Also the deposit which the FG must pay.

- FG’s pay deposit

- PMO organizes inputs from fertilizer and seed companies to stockists

- Stockists inform PMO of inputs arrival

- PMO informs FG’s to collect inputs and issues electronic token

- Stockists release inputs to FG’s, claim handling fee from PMO using same electronic token

- FG’s inform PMO dates and approximate quantities for collection (1-5 days before crop is ready)

- PMO relay that info to buyer, who schedules collections

- Buyer informs PMO the schedules who relay back to FG’s and get confirmation

- Buyer’s employee collects produce, sms’s PMO with FG , kilos, grade, and document reference

- PMO relays info to buyer, and effects immediate or almost immediate payment of 50% of crop value less loan capital, loan interest and handling fees

- Final payment to FG is effected 15 days after produce collection, by buyer through PMO.

Alex weir (alexweir1949@gmail.com)

Nairobi

11 March 2007

(working for www.prideafrica.com / www.drumnet.org , who are partnering with Equity Bank Kenya ( www.ebsafrica.com ) on this project).

Sms gateway through www.smscountry.com.

Overall project funding by various organizations including IDRC Canada (www.idrc.ca). Major software modules and components and overall software development by www.verveko.co.ke

http://whiteafrican.com/?p=414 - some press coverage

Approximate costings for operating an sfmss scheme (these are for sunflower);

50 groups, 1000 farmers, 1000 acres (400 hectares), crop is sunflower.

US$ 25/acre for fertiliser (50 kg /acre)

US$ 0.50/acre for seed (2 kg /acre)

US$ 200/acre produce (1000 kg/acre at US$ 200/tonne)

therefore if 1 acre per farmer (0.4 hectare) and 1000 farmers then approx US$ 25,500 loan requirement

gross revenue (with zero defaulters) is - 15% of loan as interest = 5 months x 3% per month (gross rate of interest approx 40% p.a.)

plus handling charge maybe 10% of loan - i.e. US$ 6,375-00 total charges to farmers (25% of gross income)

costs - bulk loan at maybe 15% p.a. net rate of interest - 5 months = US$ 1,500-00

admin - phone charges at US$ 0.04 per outgoing sms and US$ 0.02 per incoming sms -

approx 8+8 sms per group.. - and 50 groups (20 farmers per group) - US$ 25-00

admin labour – 1-2 persons @ US$ 2,000-00 per annum, US$ 1,000-00 per growing season.

computer costs - best is for me to run this as a service - there is a flat fee per group plus a fee per

transaction - these costs should be minimal - e,g, US$ 500 per annum for that kind of scale, or US$ 250-00 per growing season.

Note that these fees and charges can be modified so that for example, the scheme works as a non-profit/non-loss scheme.

Obviously the scheme becomes more attractive for operators and for farmers when the scale increases – e.g. to 1,000 groups, 20,000 farmers, and still using only 1-2 persons for labour

Other considerations when seeking partners with whom to implement such a scheme:

A bank with an affordable electronic banking system and an affordable mobile banking system.
A bank whose electronic banking system allows online payments and online statements; ideally those online payments can be effected by a payments file – i.e. bulk payments (as opposed to individual payments). Ideally those online statements are available in xml format, or at least in csv format (comma separated file). Because there is a lack of international (or even national) agreed bank payment file and statement standard formats, then usually the programming of the e-banking module itself has to be done especially (i.e. it is a bespoke module which fits into the general system). Hopefully at some time in the future standard formats are adopted globally, and the system can be written to encompass that format or formats. Note also with regard to the question of payment files, ideally these are human- and also of course machine-readable, but they also have a textfile which is submitted at the same time which acts as a checksum on the payments file itself, to eliminate the chance that the human accounts operator changes the file after approval and before submission to the e-banking system.
A bank whose mobile banking system has affordable or free sms inpayment alerts
A produce purchasing company (PPC) who are willing and are capable of organizing and implementing the transport and the produce grading and collection process.
Note that the PMO (project management organization) may be in fact part of the PPC or the Participating Bank, or may be a temporary project organization composed of a few staff from both organizations. Alternatively it may be an independent entity.
Note that there is also a requirement ideally for some kind of pre-vetting of farmers and farmers groups, so that dubious persons and dubious groups do not participate in the scheme – i.e. there is a requirement for some kind of labour on the ground. This could be NGO labour, Ministry of Agriculture labour etc etc.. The business of establishing that land has indeed been planted and that crops are growing could probably be done by using Google Earth on the internet.
There is also a requirement for various paper forms and mini-manuals. There is no need for these to be captured on the system or indeed on any computer – they can be stored in the event that any of the farmers or the groups defaults or sells the produce to another individual or organization, in which case some kind of legal action occurs or is threatened. The number of forms and manual pages has to be minimized to a manageable quantity.

Alex weir (alexweir1949@gmail.com)

Harare

17 April 2007

Postscript – Harare – 21 April 2008

I now attach to the zip file - http://www.cd3wd.com/sfmss/sfmss.zip - of this document also a zip file of the sql source code for the 95% of the system which I wrote. This is in postgressql flavour of SQL but can easily be adapted to other SQL flavours, including MSSQL (Microsoft Sql Server) if and when required. Hopefully this move will raise interest (?) in this kind of system…. If anyone wants to take the project further then contact me (http://www.cd3wd.com/contactus/ )

Best Regards

Alex weir (alexweir1949@gmail.com)

Harare