Data Analytics

By Vishnu Prasad, IFMR Finance Foundation

This post is part of our series on the current state diagnostic step.

At the end of the cadastral mapping and surveying process, our local team had produced detailed maps of physical infrastructure for every street in Srirangapatna, 256 household surveys (5% of the town’s households) and 12 unique business surveys. Our next challenge was to translate this high quality data into meaningful analysis and visually intuitive formats.

As we pondered over the various options, it occurred to us that the best way to represent our analysis was to do it spatially. Using Quantum GIS (Geographic Information Systems), an open source GIS application that is essentially MS Excel with the added dimension of spatiality, we created maps that made our analysis visually appealing and easy to comprehend.

For example, our data depicting garbage collection across the wards which looked like this:

was transformed into:

As we looked at these maps, their many advantages over conventional modes of analysis became apparent. We realized that in the map above, Ganjam (wards 16-23, on the right side) had trash collection rates that were much lower than Srirangapatna Fort Town (wards 1-15, to the left). Spatial representation of the analysis brought out many aspects that would not have been readily apparent otherwise.

We could also run analytics such as the one below:

This map shows all properties with a specified distance of the nearest streetlight. Analytics such as these became vital in determining access (proximity) to infrastructure across Srirangapatna. For example, the spatial representation of our analysis shows that slums are very often worst hit in terms of access to infrastructure.

Having created maps using Quantum GIS software, a freeware available online, we overlaid these on Google Maps. You can access all the data from our cadastral mapping on the maps section of our website.

Our household surveys gathered socio-economic data over 10 sectors: Work Habits, Transportation, Shopping, Drinking Water, Sanitation, Solid Waste, Electricity, Housing, Finance and Technology. The analysis produced reports that included comparisons across metrics between Srirangapatna Fort Town and Ganjam, and wards with and without slums.

We also produced comprehensive ward level reports like the one below which shows data for Ward No. 20.

Since the completion of the report, we have shared our results with the ward council of Srirangapatna TMC. The ease with which they understood our analysis (despite the fact that no one in our team speaks Kannada) has validated the approach that we have taken to represent our results.

We will now use these representations in our visioning exercises with citizens on the future of Srirangapatna.


Capturing business sentiment

By Dinesh Lodha, IFMR Finance Foundation

As part of our data collection efforts that involved mapping infrastructure and conducting household surveys, we also reached out to local businesses of the town to get a pulse of the local business environment.

It is said that Tipu Sultan had during his reign invited skilled craftsman and artists to settle in Srirangapatna. This clearly has had a telling effect, in the fact that the town is still known for the quality of its craftsmanship especially in relation to woodwork. I had the opportunity to meet Mr. Manjunathan, a bullock cart maker. His family has been involved in bullock cart making for generations and he had inherited this tradition from his father. It was humbling to see him treat his work more as an art than business and the level of detailing that goes into making a simple bullock cart. Like him there are around 10-12 bullock cart makers who reside nearby, a number he confesses, that has been coming down thanks to increasing mechanization. He is not sure if his children would grow up to appreciate, forget practice, this family tradition.

Like the bullock cart makers, most of the small and medium businesses are located in the Ganjam area of the town. The designated Industrial Estate houses enterprises that are into making handicrafts, wood interiors, pipes, copper wires, lead batteries, etc. The fort area of the town, being more populated, has many small retail shops especially on Pete Beedi Road that cater to the local and tourist population.

Industrial Estate

One interesting feature of the local economy is the weekly Saturday market, Santhe. Held every week on an open field behind the local bus station, the market has around 600 shops setting base every Saturday selling an array of items. It is estimated that the market has around 4000-5000 people shopping there every Saturday, with almost 60-70 percent of them being shoppers from Srirangapatna, the remaining coming from nearby villages.

The town has two large manufacturing facilities – Gokaldas Exports (owned by Blackstone, a private equity player) and MK Agrotech, each employing close to a 1000 people.

We tried to make the business surveys as representative as possible, covering the large manufacturers on one hand to small and medium businesses employing 5-10 people on the other. We also met with a couple of association presidents who amongst them represented close to 300 small retail shops in the Srirangapatna fort area. The surveys revolved around gathering basic information about the businesses, supply chain network, strategy and their relative strengths and weaknesses compared to similar businesses elsewhere.

From the conversations it was clear that most of the small and medium business owners were in the town for historic than strategic reasons. Most of the inputs for businesses came from outside and outputs were largely meant for markets outside the town. Issues around labor quality and infrequent power supply were a common concern.

The information gathered through these business surveys was largely of qualitative nature, and is designed to enable us to better appreciate the challenges that people face from an economic standpoint.


Mapping Infrastructure and Conducting Household Surveys

By Dinesh Lodha, IFMR Finance Foundation

This post is a continuation of our series of posts on the current state diagnostic step.

To get a sense of the time that it might take for our data collection efforts, we had earlier undertaken a trial data collection exercise from our end in one of the slums of the town. Based on our experience and a basic understanding of the geography, we set ourselves the target of covering the town, which is split into Srirangapatna Fort Area and Ganjam, in a span of 2 weeks.

For the duration of our data collection efforts we decided to stay in a nearby accommodation, which was close to the local municipal office. While the place had a friendly staff and an expansive menu that was largely on paper, what stood out were two large emus, who for some reason were keen on plucking our heads out every time we walked past them. Perhaps the scam around emu farming had gotten the better of them!

We chose the local municipal office as our center of operations as it was easily accessible and was easy to travel from to other parts of the town. Our local team would arrive there every morning for a briefing about the day’s schedule.

The local municipal office

We had split the group into 4 teams, each consisting of a designated surveyor and a physical mapper. The surveyor’s task was to conduct 10 household interviews per ward; with an over representation of household interviews from slums in case a ward had one. The physical mapper on the other hand would observe the parameters that we wanted to capture and draw them on the A3 sheet in line with the color codes that we laid down.

Our daily briefing would involve a combination of addressing the group as a whole and each team individually. This was to ensure that they understood the task at hand and things weren’t lost in translation, in addition to of course addressing any queries that they may not express in front of the larger group.

After the morning brief we would walk each of the teams to their respective starting locations of the ward they would be undertaking data collection for. This was primarily done so that they could correspond their present location with the map – orienting them so as to accurately map the physical infrastructure.

Local team members in action

Depending on the size of the wards and the proximity from the local TMC office, it would usually take 3-4 hours for the teams to wrap their data collection for a ward. After which they would arrive at the local TMC office, where we would sit with each team to go through the data collected by them. This was done to ensure that the data collected was in line with our expectations and it served as a way to train the teams better for their next day of data collection.

Jared cross-checking data collected by one of the teams

Throughout the course of our data collection we established a set of routines that we adhered to religiously and ensured that the local staff knew and realized their importance.

Some of these routines include:

  • Meeting at the local TMC office everyday at 10:00 AM.
  • Summarizing the work that was done the day before and addressing issues that may have cropped up.
  • Running through maps of wards that would be covered on the day and talking about any landmarks or important streets within it.
  • Allocation of wards to teams and briefing them about their particular wards, with special emphasis on slums if present.
  • Briefly explaining the legends and household survey questions.
  • Physically walking each team to the starting point of survey in their respective ward.
  • Making random visits to each team while they were doing the surveys/mapping of the ward.
  • De-brief and data validation of each team’s results at the local TMC office.

Creating and Training the Local Team

By Dinesh Lodha, IFMR Finance Foundation

The below post is a continuation of our series of posts on Current State Diagnostic step.

For our data collection efforts it was obvious that we would need to assemble a local team as they would know the terrain better and survey respondents were more likely to be comfortable dealing with them than with people from outside the town.

On one of our visits to the Kurad Beedi ward in Srirangapatna, we happened to meet Ruhi, a citizen of Srirangapatna, who had been involved in conducting surveys for some government health schemes earlier. She had an understanding of the type of skills that would be required to conduct surveys and we discussed with her the possibility of putting together a local team to conduct spatial mapping and household surveys in Srirangapatna.

Ruhi put together a team of 9 members, largely students who belonged to the Kurad Beedi Slum, one of the nine slums of the town. They were largely high school graduates, some of who were waiting to go to college, others who had started work after school and some who had even started families! Shanti, 18, was married and had a kid, while Ramanathan had discontinued his studies to help his father in making and selling clay idols outside the Ranganatha Swamy Temple. Most of the team members knew each other and were clearly excited to be part of the project, an experience that was completely new for them.

Some of the local team members with Jared

With the team in place, we realized that they needed to be adequately trained, especially considering their inexperience and the scope of work involved. Also we had to bear in mind that our collective vocabulary of Kannada consisted of barely three words – hence we had to tailor our training efforts to ensure that they clearly understood the objective, the processes and the outcomes that were expected.


Prior to assembling the local team we had conducted a trial data collection exercise in Ward 16 to get a sense of the challenges that the local team may encounter. This trial run provided us with valuable inputs on the difficulties of cadastral mapping and household surveying, which helped in structuring a day-long training program for the local team to prepare them for the rigours of data collection.

Mapping streetlights on Kurad Beedi, 1st Main Street in Ward 16

Jared & Anand on Gumbaz Road in Ward 16 during trial data collection

The training programme was split into two sessions and was designed to combine visual aids and practical field exercise so that they can apply what they have learnt during the training.

First Session: The emphasis of the first session lay in outlining the scope of the project and its importance to Srirangapatna’s future growth. We chose the first session to detail the mapping effort and how physical infrastructure needs to be mapped based on the legends that would accompany each map. Most of them were seeing a map of their city for the first time and were clearly curious about the exercise of actually filling in the maps with infrastructure data.

We shared our experience from our trial mapping effort earlier and showed how we had gone about physically mapping infrastructure and the likely things they needed to be careful about while plotting data points. This was followed by a practical field exercise where we split the team into groups of two, and went about mapping physical infrastructure in Ward 11. We wrapped the first session by reviewing maps of each group and providing feedback about their output.

A local team member plotting data points in Ward 11

Second Session: In the second session we introduced the household surveys and the objective behind them. We detailed the questionnaire and laid special emphasis on the behavioral aspects that they would need to keep in mind while interviewing respondents. The groups then undertook mock interviews amongst themselves using the questionnaire and sample answer sheets. This was followed by a review of each answer sheet and we addressed queries that they had on specific questions.

We ended the training program by summarizing the day’s learnings and set up to start the actual process from the following day.


Groundwork for Data Collection

By Dinesh Lodha, IFMR Finance Foundation

As part of the five-step process, below is the first in a series of posts on the Current State Diagnostic step.

In order to assess and finance a city’s infrastructure requirements, we need to undertake a visioning exercise to capture the collective vision that citizens have for the town. However, any meaningful conversation about a city’s future would require a basic understanding of where it stands today. Therefore, our first step was to collate data regarding the current state of infrastructure provision and service delivery in the city.

We started out by understanding the type of data that the Srirangapatna Town Municipal Corporation (TMC) already had. What we found was that data on key infrastructure which the city already had was largely out of date and what little was available was in physical format. For instance, the land use map was hand drawn from 1994, and continuous folding and re-folding of the map had made many of the features hard to distinguish.

What we quickly ascertained was that there was a need to generate good quality data, because the available information was patchy and out of date in many cases.

Consequently, our “Current State Diagnostic” process was designed to provide the necessary input for discussions with the citizens on the future of Srirangapatna. It was to be a comprehensive data-collection effort capturing granular data – on access to infrastructure, spatial distribution of infrastructure and quality of service delivery. The scope of the process encompassed all local public infrastructure, viz. public sanitation, drinking water, drainage network, waste disposal, street lights, in addition to slums and land use.

In order to capture all this information, we designed a two-step data collection process involving spatial mapping of physical infrastructure and surveying households and businesses on perceptions of service delivery. However, before this information could be collected on the ground, there was some background work to be done.

Srirangapatna is a town with a population of around 25,000, divided into 23 wards with 9 slums. We had to start out by creating a base level map of the city and ward-level maps of each of the 23 wards. The TMC had provided us a basic property-layer map of the town, upon which we drew the ward-level maps in Adobe Illustrator with relevant legends to aid the data collection process.

Srirangapatna City Map

Ward 2 Map

Legends used for physcial mapping of infrastructure (click image to expand)

The household survey was designed to capture citizen perceptions on infrastructure and services, and collected information on demographics, employment, transportation, drinking water, sanitation, solid waste disposal, electricity, housing, access to finance, and technology. The business survey on the other hand was intended to provide a perspective on the nature of the businesses, the supply-chain networks and local infrastructure.

Household survey questionnaire (Click image for PDF)

In the next post, we will discuss how we put in place a local team and the training program we devised for them.