DISTRIBUTION OF PUREIT WATER FILTER THROUGH A MICROFINANCE INSTITUTIONS MODEL IN RURAL INDIA Results from an Evaluation of Pilot Projects in Madhya Pradesh January, 2012 Prepared for: PATH Prepared by: Abt Associates Inc. This report was prepared by Ramakrishnan Ganesan and Samantha Bastian of Abt Associates for the Safe Water Project implemented by PATH with a grant from the Bill & Melinda Gates Foundation. Photo: PATH
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DISTRIBUTION OF PUREIT WATER FILTER
THROUGH A MICROFINANCE INSTITUTIONS
MODEL IN RURAL INDIA Results from an Evaluation of Pilot Projects in Madhya Pradesh
January, 2012
Prepared for:
PATH Prepared by:
Abt Associates Inc.
This report was prepared by Ramakrishnan Ganesan and Samantha Bastian of Abt Associates for
the Safe Water Project implemented by PATH with a grant from the Bill & Melinda Gates Foundation.
Photo: PATH
Abt Associates is a mission-driven, global leader in research and program implementation in the fields of health,
social and environmental policy, and international development. Known for its rigorous approach to solving complex
challenges, Abt Associates was ranked as one of the top 25 global research firms in 2010. The employee-owned
company has multiple offices in the U. S. and program offices in nearly 40 countries.
Abt Associates Inc.
4550 Montgomery Ave., Suite 800
Bethesda, Maryland 20814
USA
www.abtassociates.com
ABT ASSOCIATES INC. 1
TABLE OF CONTENTS
List of Acronyms ........................................................................... 3
1.1. Introduction ........................................................................................... 12 1.2. Pilot Partners ......................................................................................... 13 1.3. Product Promoted through the Pilot ............................................... 14 1.4. The Microfinance Institutions Model and Pilot Description ....... 15
2.1. Objectives of the Microfinance Institutions Model ....................... 20 2.2. Evaluation Methods .............................................................................. 20
2.2.1. Baseline and Endline Household Research ....................... 21
2.2.2. Qualitative Research Study .................................................. 21
2.2.3. Longitudinal study .................................................................. 22
2.2.4. Collation and Analyses of Project Data, and Interviews
with Project Partners .................................................................... 23
3. Impact on Purchase and Use of PureIt ......................... 25
3.1. Respondent Characteristics ............................................................... 25 3.2. Source of Drinking Water .................................................................. 26 3.3. Water Quality Perceptions ................................................................ 27 3.4. Perceived Need for Water Treatment ........................................... 28 3.5. Potential Market for PureIt ................................................................ 29 3.6. Exposure to Project Activities ........................................................... 31 3.7. Perception of PureIt ............................................................................. 32 3.8. Awareness, Purchase and Use of PureIt ......................................... 34 3.9. Determinants of Purchase of PureIt ................................................. 36 3.10. Determinants of Continued Use of PureIt .............................. 40 3.11. Correct and Consistent Use ....................................................... 43
3.11.1. Correct and Consistent Use in Nagda ........................... 44
3.11.2. Correct and Consistent Use in Neemuch ..................... 47
4. Assessment of Commercial Viability of the Model ...... 50
4.1. Approach to Assessing Commercial Viability ................................ 50 4.2. Market Potential and Impact of Marketing Activities ................... 50
4.2.1. Design Changes During Implementation .......................... 50
ABT ASSOCIATES INC. 2
4.2.2. Increase in Awareness and Purchase of PureIt ............... 51
4.2.3. Effectiveness of Marketing Activities ................................. 51
4.2.4. Continued Use of PureIt ...................................................... 52
4.3. Analyses of Sales and Revenues ........................................................ 52 4.3.1. Sales Trends ............................................................................. 52
4.3.2. Level of Commercial Viability ............................................. 53
4.4. Partners‟ Willingness to Continue with the Model ...................... 58 4.4.1. HUL‟s interest in continuing the model ............................ 58
4.4.2. Spandana‟s Interest in Continuing with the Model ........ 58
4.5. Model Results and Potential: Summary ........................................... 59
5. Limitations of the Research ........................................... 62
Many individuals contributed to the design and implementation of the study, provided feedback on the
presentation of the study results, or contributed valuable comments on an earlier draft of this report.
The authors would like to acknowledge the contributions of Slavea Chankova, Brad Lucas, Francoise
Armand, Sidhartha Vermani, Tanya Dargan, Lorelei Goodyear, Elizabeth Blanton, Siri Wood, Tim Elliott,
Jody Garcia, and Claudia Harner-Jay.
ABT ASSOCIATES INC. 7
EXECUTIVE SUMMARY
Introduction
PATH's Safe Water Project (SWP) aims to assess gaps in the household water treatment and safe
storage market and to determine how well private sector companies can successfully and sustainably
reach lower-income consumers with effective household water treatment and safe storage products.
The Microfinance Institution (MFI) model is one of a number of pilot projects that PATH and partners
undertook in India to find ways of overcoming distribution and marketing barriers that make it difficult
for manufacturers of household water treatment and safe storage products to penetrate lower-income
markets. The model was piloted in three Indian states: Tamil Nadu, Andhra Pradesh and Madhya
Pradesh. In this model, durable water filters were sold to rural and peri-urban consumers through the
MFI networks. The model works by using the MFI network to (i) aggregate demand for a durable high-
value product and (ii) by providing the product to low-income customers at affordable installments
payment schemes. In Madhya Pradesh, India, the MFI model was piloted through a partnership with
Hindustan Unilever Ltd (HUL), the marketer of PureIt, a durable water filter, and Spandana
Microfinance, a large microfinance institution in India. PureIt was sold at a subsidized price to MFI clients
who took out Spandana loans for the purchase.
Objectives of the Pilot
The MFI model sought to test whether promoting and distributing PureIt through Spandana‟s loan group
meetings and network to the MFI‟s clients can be sustainable, and the extent to which such a model can
create increases in correct and continued use of PureIt among low- and middle-income consumers.
The key questions for the MFI model are:
1. What is the PureIt uptake rate among target consumers?
2. What is the extent of correct use of PureIt among target consumers?
3. What are the triggers and barriers to the uptake and continued use of PureIt?
4. Can commercial partners earn a profit from sales to target consumers?
5. Will commercial partners continue and/or scale up the SWP pilot business model to reach
target consumers?
6. What other efforts/inputs/incentives are needed to stimulate supply and demand for PureIt to
target consumers?
In addition, the pilots in Madhya Pradesh build on the knowledge gained and lessons learned from the
earlier MFI pilots and seek to inform the following variant-specific questions:
Are the uptake levels achieved in new markets broadly comparable to those achieved in established
markets?
Does product cost-defrayment increase levels of uptake and consistent use?
Can the cost-defrayment be recouped through sales of the replacement germ-kill kit?
Pilot Description
In Madhya Pradesh, the MFI model was implemented in two Spandana branches (Nagda and Neemuch)
with approximately 5,650 clients. In Nagda, the intervention was implemented over five months
ABT ASSOCIATES INC. 8
(February-July 2010). After the pilot in Nagda, the intervention was rolled out in Neemuch. However,
the intervention was curtailed due to the nationwide MFI crisis that took place at the same time.
Therefore, the intervention was only implemented for a month in Neemuch (September 2010).
The pilot intervention involved the sale of PureIt water filters through the Spandana network. PureIt is a
durable water filter that employs a combination of filtering techniques – a net filter (to remove visible /
coarse impurities), chlorine (automatically released through a cartridge, - microbicide), activated
carbon/charcoal filter (to remove chlorine smell/taste). The replaceable part of PureIt is known as the
Germ Kill Kit (GKK).
The pilots in Neemuch and Nagda differed in the amount of subsidy and the loan installments offered to
the MFI clients. In Neemuch, PureIt was sold at a 50-percent cost defrayment. Customers paid Rs 1,000
in loan installments of Rs 43 over 25 weeks. In Nagda, there was a 100-percent cost defrayment scheme
for PureIt, and customers who received the free filter had to pay Rs 730 for two replacement GKKs
(packaged with the device) with a loan installment of Rs 39 for the first 17 weeks and Rs 37 for the last
three weeks.
The promotion and distribution of PureIt involved the HUL representative using the platform of the
Spandana group meeting to make presentations of the need and methods to treat water, highlighting
PureIt as a promoted water treatment method. Following the meeting, if the MFI client decided to
purchase PureIt, the loan was sanctioned by the Spandana loan officer and the PureIt water filter was
distributed by the HUL representative.
The resupply of the GKK was done by Spandana while the after sales services were provided by HUL.
Evaluation Approach
The key research questions and additional related topics considered in evaluating this pilot were
addressed through a mix of primary and secondary data sources.
These sources included:
Baseline and endline household research: The baseline and endline household surveys were designed to
assess the extent of increases in uptake and regular use (reported) of PureIt, as well as changes in
other proximal indicators of interest to the project. To answer these questions, we carried out a
pre-post study in the intervention area, using data from a population-based survey of households.
The baseline survey for both Neemuch and Nagda was conducted in December 2009 (before the
start of the intervention) and the endline survey for Nagda was conducted in November 2010 and
for Neemuch in January 2011. A sample of 440 and 414 households were drawn at baseline and
endline respectively in Neemuch. In Nagda, a sample of 440 and 402 households were drawn.
Qualitative research study: The purpose of the qualitative research study was to understand motives
and barriers to trial and continued use of PureIt (research question no. 3). This study was
conducted in June 2011. A total of two focus group discussions and six in-depth interviews were
conducted among women (and their spouses) in the intervention area who was a MFI client.
Longitudinal research: After the endline household survey, a longitudinal survey was conducted with
follow-up visits for six months among those respondents who purchased PureIt and were using it at
endline. The longitudinal study aimed to measure the rates of correct and consistent use of PureIt
over time, and provide data on the reasons for discontinuing use of the filter.
Collation and analyses of project data, and interviews with project partners: We utilized project cost,
sales, and revenue data. We interviewed the project partners to understand how the project was
implemented, whether it met their expectations, the challenges encountered and the lessons
ABT ASSOCIATES INC. 9
learned. These, along with survey research data, were used to answer research questions related to
commercial viability.
Findings
The results from the household surveys and the qualitative research study clearly show that:
Felt need was high; however this was mainly due to visible impurities in the water and did not arise
from the need for microbial safety. Respondents therefore engaged in filtering with cloth/plastic
sieve and using alum to treat their water.
At baseline, the intention to buy PureIt was low for the 50-percent cost defrayment scheme. The
key segment identified as potential purchasers was those who were placed in the upper wealth
quintile and boiled their water. However, even within this segment, the intention to buy at 50
percent cost defrayment was below experiential norms developed in the SWP project. The potential
market for the 100-percent cost defrayment scheme was not evaluated at baseline.
Findings from the household surveys and the qualitative research study show that:
o Awareness of PureIt increased significantly from 4 percent and 7 percent to 56 percent and
74 percent in Neemuch and Nagda respectively.
o There was a significant increase in purchase of PureIt in both pilots. In Nagda where the 100
percent cost-defrayment model was implemented the uptake was an impressive 44 percent.
However, the purchase level was much lower in Neemuch (7 percent).
o 72 percent of those who purchased were from the SWP target group (middle three wealth
quintiles).
o The high uptake rate reported in Nagda is offset by qualitative research findings that
profiting from resale may have been a strong reason for purchase. As many as 35 percent
did reported having given away the PureIt they purchased. The quantitative and qualitative
surveys indicated a high level of re-sale of PureIt. Further, the qualitative research also
indicated that some Spandana clients may have purchased PureIt for other non-member
households belonging to higher wealth quintiles.
The research findings suggest five important reasons for not purchasing PureIt:
o Low exposure to promotional activities and scheme curtailing, particularly in Neemuch.
o Low of perceived need for water treatment (microbial safety)
o Poor perceptions of PureIt in terms of value for money and product efficacy
o Lower perceived self-efficacy and family support
o Lower affordability, even with cost-defrayment
In Neemuch activities were curtailed due to the MFI crisis. Due to this, the levels of recall of
activities (both coverage and intensity/frequency of interactions) in Neemuch were substantially
lower than in Nagda. Consequently, we cannot assess the effectiveness of the Neemuch pilot where
the 50-percent cost defrayment model was implemented.
Impact of intervention on uptake suggests that Spandana meetings alone significantly increase the
chances of purchase. Additional media exposure increase possibility of purchase but between
Spandana meetings and both mass media and Spandana meetings there is no significant increase
These studies report important reasons for the high lapse rates, in particular:
ABT ASSOCIATES INC. 10
o The reasons for lapsing were similar to the reasons for not purchasing PureIt in the first
place. The low value perception of PureIt, along with the lack of irreplaceable benefits of the
product, translated to gifting and reselling PureIt.
o Replacement of GKK was not a reason for lapsing; however, price and availability of GKK
could lead to further increase in lapse rate in the future.
o From the longitudinal study, we found that in both pilot areas current use dropped over the
six to seven months of follow-up. Correct use was irregular due to intermittent
maintenance by the respondents, resulting in negligible rates of consistent correct use by the
last track of the longitudinal.
Assessment of the effectiveness of marketing activities leads to the hypothesis that the cost-
defrayment may have resulted in a negative externality – dis-incentivizing HUL product
representatives in consumer education and motivation.
Without the cost-defrayment, HUL‟s gross margins are likely to partially or fully cover all costs
directly attributable to this model. Thus, without cost defrayment, the model would be at a cost
recovery level 4 or 5. However, with cost-defrayment, we classify the level of commercial
sustainability (or cost recovery) as level 3 – it recovers product and distribution costs, but does not
fully recover marketing costs deployed towards cost-defrayment.
The net loss to HUL from sale of a PureIt in Neemuch is Rs 630, and the net loss in Nagda where
PureIt was bundled with two additional GKKs for Rs 730 is Rs 1,322. To recover these losses from
sales of GKK, the average number of additional GKKs purchased by those who purchased PureIt in
Neemuch needs to be 4.5 or more. Similarly, the average number of additional GKKs purchased by
those who purchased PureIt in Nagda needs to be 9.5 or more. The continuation rates observed in
the longitudinal survey show that such levels of GKK re-purchase are unlikely.
Thus, the commercial viability for the different partners in this model are summarized as:
o High level of commercial viability for Spandana, with product, distribution and marketing
costs, and overheads and profit components nearly fully recovered.
o Full commercial viability for HUL‟s distributor who recovered completely all of these cost
components.
o Incomplete cost-recovery for HUL, with production and distribution costs recovered
but not marketing costs (cost-defrayment), overheads and profits.
In the Neemuch pilot, total cost recovery was 93%. HUL achieved 72% cost recovery, while
Spandana achieved a 122% cost recovery.
In the Nagda pilot, total cost recovery was 85%. Cost recovery rates were 62% for HUL, and 130%
for Spandana.
Interviews with HUL indicate that HUL will not be willing to continue with the cost-defrayment
model. However, HUL reported that it was keen to continue and scale-up the MFI partnership
approach without cost defrayment. Spandana‟s interest in continuing the model could not be
assessed since Spandana representatives were unavailable for interviews.
Recommendations
The findings point to the following recommendations on changes to the model design and
implementation that partners should consider to improve the health impact and commercial viability of
ABT ASSOCIATES INC. 11
this model. Some of these recommendations are known to be feasible and effective on the basis of
HUL‟s experiences with this and similar pilots.
Increased intensity of coverage: The findings suggest that there is a need for increased intensity of
activity both in terms of quality and quantity of group meetings conducted. There is a need to
involve the entire family especially the spouse and elders in the decision making process in order to
ensure uptake and continued use. HUL has seen benefit in greater involvement of the MFI CAs and
NGO staff in conducting repeat visits to households.
o Spandana meetings alone provide a significant positive impact toward uptake rates.
Resources could be focused on MFI group meetings to increase uptake of PureIt. This is
essential in converting latent need for water treatment to demand for PureIt.
Cost defrayment is an ineffective route: those who did not intend to use PureIt also purchased
leading to a mis-targeted subsidy. The model was therefore successful in generating high purchase
rates; however, it also saw a high incidence of lapse and resale. Uptake and continued use is likely to
be impacted positively if the cost defrayment resources are used towards increasing intensity of
activities and frequency of interaction rather than to substitute product price.
Price and availability of GKK may come up as a barrier to continued use in the future. There is
therefore a need for a feasible GKK supply chain and product bundle. HUL has had positive
experiences with the sale of two LLBs for Rs 1,000 repaid over 50 weeks.
Some product related functional barriers are heightened and may take a more central position as a
cause for lapse once the other concerns are resolved. These include the problems faced by large
families that require larger volume of water - frequent refilling and replacement of GKK may be
cumbersome – and the delicate appearance of PureIt and difficulties in finding a suitable place for it
make it less desirable in smaller houses. Resolving these issues may be challenging as it likely needs
to involve creating different versions (in terms of size and sturdiness) of the device to address the
needs of these user segments.
ABT ASSOCIATES INC. 12
1. BACKGROUND
1.1. INTRODUCTION
The Safe Water Project (SWP), implemented by PATH with a grant from the Bill and Melinda Gates
Foundation, aims to develop sustainable distribution models for ensuring access to Household Water
Treatment and Safe Storage (HWTS) products, together with promoting sustained use of those
products over time by low-income populations. Commercial model pilots are being implemented in a
phased manner to assess and refine various HWTS products and business approaches. The primary
objective of the pilots is to assess various distribution models for reaching lower-income populations
with products that are already commercially available and effective. Distribution models and HWTS
products piloted by the SWP include:
Bicycle entrepreneur sales model for chlorine tablets (Aquatabs) in rural Uttar Pradesh, India
Microfinance institutions (MFI) model for table-top combination filters in Tamil Nadu, Madhya
Pradesh and Andhra Pradesh, India
MFI model for ceramic water pot filters in Kampong Speu, Cambodia
Direct (door-to-door) sales model for ceramic water pot filters in Kampong Speu, Cambodia
Unorganized retail distribution model for ceramic water pot filters in Kampong Spam, Cambodia
Basket-of-goods model for ceramic water pot filters in Nyanza province, Kenya
Promotion and distribution of Aquatabs by community health workers in Kan Tho province,
Vietnam
The MFI model seeks to test whether an MFI network can be leveraged - as a demand aggregator and
financier - to promote and distribute a durable water filter to lower-income households sustainably; and
the extent to which such a model can increase uptake and correct and consistent use of the water filter.
Variants of the MFI model were piloted in Tamil Nadu, Madhya Pradesh and Andhra Pradesh. In Tamil
Nadu and Madhya Pradesh the pilots were implemented in partnership with Hindustan Unilever Ltd
(HUL), the manufacturer and marketer of PureIt and Spandana Sphoorty Financial Limited (Spandana),
one of the largest MFIs in India. A key factor in the MFI model variants that were tested in Madhya
Pradesh was the effect of cost-defrayment (i.e. sale of the water filter at lower than retail price) on
uptake and continued use of the product.
Abt Associates lead the monitoring and evaluation (M&E) of these pilots. The M&E approach included
household surveys to assess increases in uptake and continued use of PureIt, identifying motives and
barriers to purchase and continued use through exploratory research studies, longitudinal surveys to
measure correct and continued use, and assessment of the commercial viability of the model using sales
and cost data and information from pilot project partners.
This report presents the evaluation findings for the MFI model in Madhya Pradesh. Separate evaluation
reports were produced by Abt Associates for each of the other SWP pilots.
ABT ASSOCIATES INC. 13
1.2. PILOT PARTNERS
PATH‟s partners for this pilot – HUL and Spandana – are described below.
Hindustan Unilever Limited (HUL) is a subsidiary of Unilever, one of the world‟s leading suppliers
of fast moving consumer goods with strong local roots in more than 100 countries across the globe and
annual sales of about €44 billion in 2011. In India, HUL has over 35 brands spanning 20 distinct
categories such as soaps, detergents, shampoos, skin care, toothpastes, deodorants, cosmetics, tea,
coffee, packaged foods, ice cream, and water purifiers. HUL is renowned for its marketing expertise and
extensive distribution network. Six HUL brands were listed among the top 10 most trusted brands in a
survey in 2010 (Brand Equity's „India‟s Most Trusted Brands Survey‟ rankings for 2010). HUL directly
reaches over one million retail outlets and its products reach over 6.3 million retail outlets.
HUL entered the water purification business in 2004 with the introduction of PureIt - a combination
gravity water filter. Research and development for PureIt was driven primarily from HUL‟s R&D center
in Bangalore – one of Unilever‟s six principal R&D centers globally. The product is now slated for
introduction in other countries too. 1 PureIt was initially introduced in Tamil Nadu through a retail
distribution model, and the bulk of HUL‟s activities including intensive multi-media brand promotion
campaigns were focused in this state until 2008-09.
HUL gradually expanded its coverage of PureIt to other states, and coverage in Madhya Pradesh began in
2009. Another key aspect of HUL‟s strategy for PureIt from 2008-09 onwards was the emphasis placed
on „direct sales‟, whereby HUL began developing and strengthening internal resources and personnel
towards this model. Parallel to this, HUL entered into partnerships with NGOs where PureIt was jointly
marketed as a means to increase use of water treatment solutions in the community. HUL characterizes
its partnerships with NGOs prior to the SWP project as „individual-driven‟ and „very few significant
partnerships‟ (i.e. partnerships with MFIs and NGOs were not a part of its core business strategy for
PureIt).
HUL‟s water purification product business had not achieved break-even in 2009; it is now expected by
HUL to break-even by 2012. With increased emphasis on direct sales, HUL was interested in exploring
direct sales models based on social partnerships with an aim to expand its target market from socio-
economic class (SEC) A initially to include SEC B, C and D households.2
Spandana Sphoorty Financial Limited is the second largest microfinance institution in India. It was
started as an NGO in India in 1998. By 2010 Spandana was present in 10 states, and had 4.2 million
clients. Initially supported by Friends of Women World Bank (FWWB), Spandana later also received
funds from development financial institutions, public and private sector banks, including foreign banks,
and other non-bank financial corporations.
The MFI lends to low and lower-middle income households typically borrowing for investments in
agriculture, dairy and micro-enterprises. It also provides services to improve the livelihoods of its low-
income clients. These include initiatives for safe drinking water, renewable energy and insurance. The
microfinance institutions lend to groups of women at group level and individually using the Grameen-
style Joint Liability Group (JLGs) model.
1 HUL has invested in other product line extensions (new HWTS products using the Pureit brand name). PureIt
Compact, for example, is a Rs. 1,000 HWTS product developed by HUL. There are also more expensive products. 2 SEC is a classification of consumers developed by India‟s Market Research Users Council. In urban areas, A is the highest
SEC, followed by B, C, D, and E. Box 1 later in the report provides a more detailed explanation of how SEC segments are
defined.
ABT ASSOCIATES INC. 14
Spandana had prior experience in selling products that improve health or quality of life to its customers.
However, most of its interactions with manufacturers or marketers were limited to „bulk purchase‟ -
purchasing from manufacturers or marketers at wholesale prices and distributing the product to their
clients via loan installment payments.
In October 2010, all microfinance institutions in India were placed under scrutiny due to tensions
between the government-sponsored self-help group model, and the privately-funded microfinance
institutions (MFIs) with JLGs3. While this crisis was precipitated by adverse media coverage of the initial
public offering of one of the largest MFIs (SKS microfinance), it affected operations of all MFIs based in
Andhra Pradesh, including Spandana, and brought most of the MFI‟s operations to a stand-still.
1.3. PRODUCT PROMOTED THROUGH THE PILOT
PureIt, the product marketed in this pilot is a durable combination gravity filter. It does not require
electricity or continuous tap water connection. The total capacity of the filter is 23 liters, of which the
purified water storage capacity is 9 liters.
It purifies water using a four stage filtration mechanism (Figure 1) that consists of: (i) Microfibre Mesh™
- removes visible dirt; (ii) Compact Carbon Trap™ - removes harmful parasites and pesticides; (iii)
Germkill Processor™ - uses 'programmed chlorine release technology' to target and remove invisible
harmful viruses and bacteria; and (iv) Polisher™ - removes chlorine and ot her contaminants to make
water clear, odorless and improves taste.
The Compact Carbon Trap™, Germkill Processor™ and Polisher™ need to be replaced every 3 to 6
months, based on usage (i.e. liters of water purified). This
consumable is sold by HUL as a Germkill Kit™ (GKK).
PureIt has an indicator, which alerts the customer a few days in
advance (based on average use) when the GKK needs to be
replaced. If the GKK is not replaced on time, PureIt has an „Auto
Shut Off‟ mechanism that automatically stops the water flow until
the GKK is replaced.
3 Intellecap, White Paper, „Indian Microfinance Crisis of 2010: Turf War or a Battle of Intentions?‟ -
1 Source of data is NFHS. 2 Source of data is the SWP survey.
3.2. SOURCE OF DRINKING WATER
The surveys showed that the main source of drinking water in the pilot areas was piped water, primarily
from public taps/stand pipes (Table 8). As per the baseline study, there were no significant differences in
the sources of drinking water accessed in the dry and rainy season. However, there were significant
changes from baseline to endline9 in the types of source of water used by the study population– in
8 The NFHS is the India equivalent of the more widely known Demographic and Health Survey (DHS). 9 The baseline and endline surveys were conducted about a year apart in both pilot areas (so in the same season).
ABT ASSOCIATES INC. 27
particular an increase in piped water into yard/plot as the main source of drinking water in Neemuch
and an increase in public tap/standpipes as the main source of drinking water in Nagda.
Table 8: Drinking Water Source (dry season)
(percent of respondents)
Neemuch Nagda
Baseline Endline Sig.1 Baseline Endline Sig.1
N=440 N=414 N=440 N=402
Piped into dwelling 10.2 8.1 14.1 13.7
Piped into yard/plot 11.2 33.6 * 5.9 6.0
Public tap/standpipe 64.7 28.9 * 61.0 69.2 *
Tube well/borehole 1.1 18.0 * 10.0 6.2 *
Other 12.8 11.5 9.0 4.9 * Notes:
1Significance of difference between baseline and endline marked with * refers to p<0.05
Proportions reported are adjusted for socio-demographic variables.
The qualitative research carried out in June 2011 provided an understanding of the changes observed in
drinking water sources between the baseline and endline quantitative surveys. In both areas, qualitative
research respondents reported recent improvements in water supply infrastructure and in the quality of
water from these sources. The qualitative research study also found that in Neemuch the ground water
was supplied by the government through pipelines; and in Nagda, water from the Chambal River was
piped by the government, as well as through an initiative of Birla Gram (the main industry in Nagda).
3.3. WATER QUALITY PERCEPTIONS
Perceived quality of drinking water from the main source was measured in the surveys using a 3-point
semantic differential scale (excellent, average, bad). Table 9 shows that most of the respondents
perceived the quality of their drinking water to be average or bad. The changes in perception of water
quality can likely be explained by the changes in water sources available in the area, as well as potential
increase in awareness regarding water quality that may have been generated through the pilot.
Table 9: Water Quality Perception
(percent of respondents)
Neemuch Nagda
Baseline Endline Sig.1 Baseline Endline Sig.1
N=440 N=414 N=440 N=402
Opinion for the source of drinking water in dry season
Excellent 21.8 5.8 * 12.6 8.4
Average / so-so 61.7 90.5 * 52.4 91.8 *
Bad 16.5 3.7 * 35.1 0.0 *
Notes: 1Significance of difference between baseline and endline marked with * refers to p<0.05.
Proportions reported are adjusted for socio-demographic variables.
ABT ASSOCIATES INC. 28
The qualitative study showed that water was considered to be safe if it had no smell, no discoloration, no
visible impurities and tasted good or better than in the past. There was an absence of true knowledge of
germs. This suggests that perceived quality of water seemed to be driven by turbidity rather than microbial
aspects.
3.4. PERCEIVED NEED FOR WATER TREATMENT
The survey captured a direct measure of the extent of perceived need for water treatment in the
intervention areas. Respondents were asked how important they thought it was to treat the water they
get before drinking and responses were coded on a 4-point semantic differential scale. This scale rated a
perceived need based on: absolutely necessary, good to do when possible, not important, and not at all
required. Table 10 shows the importance of water treatment. Most reported that water treatment was
„absolutely necessary, no matter what the situation‟ or „good to do whenever possible‟. The felt need for
treating water was also very high. Felt need is defined as those who reported that their source of water
was average or bad and also reported that it was absolutely necessary / good to treat drinking water
collected from that source.
Table 10: Importance for Water Treatment
(percent of respondents)
Neemuch Nagda
Baseline Endline Sig.1 Baseline Endline Sig.1
N=440 N=414 N=440 N=402
Importance of the water treatment in dry season
Absolutely necessary, no matter what
the situation
60.4 77.6 * 32.3 87.8 *
Good to do whenever possible 34.0 21.6 * 51.5 10.7 *
Not so important – does not matter if
you do not
4.4 0.2 * 14.2 1.3 *
Not at all required - it would be a
waste
1.2 0.6 1.9 0.2
Feel need to treat water in dry season 73.1 94.2 * 75.2 90.9 *
Notes: 1Significance of difference between baseline and endline marked with * refers to p<0.05.
Proportions reported are adjusted for socio-demographic variables.
Results from the qualitative research study showed that, although there was a high felt need to treat
drinking water, the perceived need for treatment stemmed from removing visible impurities which was
typically addressed by filtering with a cloth/plastic sieve or by settling and decanting (often using alum as
a coagulant). Microbial treatment of water was generally not thought to be necessary. Boiling was
practiced as a curative measure at times of illness (as it is often recommended by doctors), rather than
as a preventive method.
ABT ASSOCIATES INC. 29
3.5. POTENTIAL MARKET FOR PUREIT
Information from the baseline survey was used to assess the potential market for PureIt in the pilot
areas. Specifically, three factors were considered: i) methods used currently to treat drinking water, ii)
awareness and use of PureIt for drinking water treatment, and iii) intention to buy PureIt as a method to
treat drinking water.
Table 11 shows the reported main method used regularly (every day or almost every day) by
households in the intervention area. Over 80 percent of the respondents filtered through a cloth/sieve
or decanted their water to remove visible impurities; 6 percent reported using boiling, and another 6
percent used chlorine, iodine or alum (among these most used alum). Use of advanced durable filters at
baseline was negligible.
Table 11: Method Used Everyday / Almost Everyday to Treat Water at Baseline
(percent of respondents)
Neemuch Nagda
N=440 N=440
Do nothing 3.3 3.2
Use cloth/net/sieve filter or settle & decant 85.4 80.6
Use boiling 6.0 6.5
Use chlorine/iodine/alum based products 4.5 8.4
Use durable filters 0.8 1.3 Note: All proportions are adjusted for variations in socio-demographic variables.
The baseline survey showed that total awareness of PureIt was very low: 7 percent in Neemuch and 9
percent in Nagda. Further, considering those reporting boiling their water regularly as the likely early
adopters suggested that the potential market for PureIt at baseline was very small – approximately 6
percent.
All respondents in the baseline survey were shown a concept card describing PureIt (product features,
benefits, how to use and price). They were then asked their intention to buy PureIt if the product was
introduced in their community at four variations of loan installments: (i) Rs 80*25 weeks (ii) Rs 60*35
weeks (iii) Rs 43*50 weeks or (iv) Rs 43*25 weeks (a 50-percent cost defrayment scheme). Figure 3
shows the reported intention to try PureIt at baseline in Neemuch and Nagda at each scheme.
Overall, reported intention to buy PureIt at full price with repayment over 25 weeks, 35 weeks and 50
weeks was low in both branches - only about 30 percent reported an intention to buy at full price with
any repayment plan. Reported intention to buy PureIt was (expectedly) highest at 50-percent cost
defrayment (Rs 43 x 25 weeks). However, it should be noted that the reported intention to buy at
baseline did not meet or exceed experiential norms of 80 percent even with 50 percent cost-
defrayment indicating likelihood of low uptake.
ABT ASSOCIATES INC. 30
Figure 3: Intention to Try PureIt at Baseline
We examined the reported intention to buy PureIt at 50-percent cost defrayment among different
consumer segments to identify sub-groups where the intention to buy meets or exceeds the experiential
norms set for this study (Figure 4). These analyses showed that the reported intention to buy was
highest among those who reported boiling their drinking water currently and belonged to upper wealth
quintiles. We interpret the findings as reinforcing the hypothesis that the potential market for PureIt
among the target population at baseline was small. Further, we interpret these findings to indicate that
felt need and ability to pay are likely to be key barriers to uptake and use of PureIt.
Figure 4: Intention to Try PureIt at 50-percent Cost-Defrayment at Baseline,
by Respondent Category
6 9 6
20
22 21
21
22
0
20
40
60
80
80x25 60x35 43x50 43x25
% r
esp
on
den
ts (
wei
ghte
d)
Neemuch
Somewhat likely to buy Very likely to buy
4 4 6
29 20 26
20
29
0
20
40
60
80
80x25 60x35 43x50 43x25%
res
po
nd
ents
(w
eigh
ted
)
Nagda
Somewhat likely to buy Very likely to buy
50 46
57 58
71
0
20
40
60
80
All Use boilingregularly
Useproduct/durable
regularly
Belong to upperwealth stratum
Boil and inupper wealth
stratum
%re
spendents
ABT ASSOCIATES INC. 31
From the analyses of the market potential for PureIt at baseline, there was strong evidence that
extensive demand-generation activities aimed at stimulating the need for water treatment beyond using
cloth/plastic filter and alum would be required to promote uptake of PureIt under the 50-percent cost
defrayment scheme. Another limiting factor for the market potential that emerged was ability to pay. It
was unclear, at the time of the baseline, how the 100-percent cost defrayment scheme would be rolled
out. Therefore the intention to buy and the potential market under the 100-percent cost defrayment
scheme was not determined at baseline.
An additional challenge confronting the pilot was the low awareness and use of any durable filter and the
low awareness of PureIt in particular. This implied that the target audience was not familiar with the
durable water filter category and brands of durable water filters. Thus, along with addressing the low
level of felt need and low ability to pay, the pilot needed to improve the target audience‟s awareness and
perceptions of durable water filters (in particular PureIt) as a needed and cost-effective solution to
water treatment.
3.6. EXPOSURE TO PROJECT ACTIVITIES
Overall, 56% in Neemuch and 74% in Nagda were aware of PureIt at endline, which was a substantial
increase from baseline in both areas.
At endline, respondents were asked whether they had seen or heard any messages on PureIt in the
previous six months. The main source of messages on PureIt reported by the respondents was TV
followed by HUL /Spandana representatives (Table 12). In Neemuch, the level of recall of messages on
PureIt from HUL/Spandana representatives was lower than in Nagda. As mentioned earlier, the pilot in
Neemuch was curtailed at one month, whereas the pilot in Nagda continued for six months. This
difference in pilot duration is likely reflected in the differences in the levels of recall of messages on
PureIt between Neemuch and Nagda.
Table 12: Sources of Messages on PureIt at Endline
(percent of respondents who recalled hearing/seeing a message on PureIt)
Main Message sources Neemuch Nagda
TV 87 82
HUL/Spandana rep 43 81
Other PureIt users 10 1
Doctor 5 9
Friends/neighbors 31 17
Table 13 summarizes the level of exposure to different types of messages and promotion activities on
PureIt, as recalled by those who had seen or heard about PureIt in the six months prior to the endline
survey. Exposure to each of the five key activities conducted during the Spandana group meetings in
relation to the pilot was substantially higher in Nagda than in Neemuch.
ABT ASSOCIATES INC. 32
Table 13: Type of Message on PureIt Recalled at Endline
(percent of respondents who recalled hearing/seeing a message on PureIt)
Indicators of Exposure to Intervention Neemuch Nagda
Heard a PureIt / HUL representative describing the importance of treating water at
home
47.7 64.5
Seen a demonstration of how PureIt works 39.2 63.3
Heard about the features and benefits of PureIt 44.3 64.5
Heard about the cost and repayment scheme 37.5 64.3
Heard about parts of the PureIt water filter that may need replacement 29.7 63.3
In addition, the pattern of intervention activities from Spandana group meetings recalled in Neemuch and
Nagda differed, reflecting different intensity of promotion activities in the two areas. In Nagda, nearly all the
respondents who recalled any activity recalled all five key activities conducted during the group meetings.
However, in Neemuch there were variations in the level of recall of individual activities conducted in the
group meetings; specifically, the recall of PureIt demonstrations and discussion on parts of PureIt that may
need replacement was much lower than in Nagda.
The qualitative research study suggests that many MFI clients in both branches had been exposed to PureIt
advertisements on TV. The qualitative research further highlighted differences in the nature of recall of PureIt
promotional activities conducted in Neemuch and Nagda. In Nagda, all respondents interviewed in the
qualitative research phase recalled 3 - 4 meetings and demonstrations. These respondents did not recall any
focus on the marketing efforts towards households with children under the age of 5 years. In Neemuch, there
was a difference in the number of activities recalled by those who had purchased PureIt vis-à-vis those who
were aware of PureIt but had not purchased it: those who purchased PureIt recalled 3 - 4 group meetings
where PureIt was discussed and demonstrated, while those who were aware of but had not purchased PureIt
recalled only one meeting where PureIt was discussed and reported that they were not shown the product
demonstration.
Thus, in addition to the shorter pilot duration, Neemuch appears to have received less coverage and likely
lower quality/intensity of promotional activities as compared to Nagda.
3.7. PERCEPTION OF PUREIT
A key insight gained from the qualitative research was respondents‟ understanding and perception of
PureIt. In both Neemuch and Nagda, respondents referred to PureIt as a „filter machine‟ that was a
coarse filtration device (channi). Interestingly, respondents did not refer to PureIt by its brand name.
The qualitative study showed that the spontaneous association to PureIt was that of a modern looking
device that removes visible impurities and makes water look clean. However PureIt had some negative
perceived attributes, including; (i) small storage capacity rendering it inappropriate for large families, (ii)
it looks fragile and delicate and may break if not kept safely (which is challenging in a small house), and
(iii) it keeps water warm and is therefore not desirable in the summer.
In both the pilot areas, very few respondents recalled explanations of germs and the role and
importance of PureIt as a „germ kill‟ water treatment method. Thus, the brand (PureIt) and its advantage
over other filtration techniques of ensuring microbiological safety were not clearly established in either
of the pilot areas, and PureIt was perceived primarily as a modern and attractive coarse filtration device.
ABT ASSOCIATES INC. 33
The baseline and endline survey findings reflect the findings from the qualitative research regarding the
perception of PureIt. The surveys captured respondents‟ perceptions of PureIt on a number of
attributes using a 3-point semantic differential scale anchored by “completely agree” and “strongly
disagree” with each statement. Table 14 shows the mean scores at baseline and endline among
respondents who were aware of PureIt. The table highlights three important factors regarding changes
in perception of PureIt among the target audience in Neemuch and Nagda:
In Neemuch, the perception of PureIt is worse-off at endline as compared to baseline (though the
difference is not significant for all items). This is an expected result where people who have heard of
the product/brand recently have not internalized its specific features and benefits. Thus, though the
level of awareness of PureIt has increased, there is a more diluted perception of the product.
In Nagda, the perception of PureIt at endline has improved over baseline. However, the
improvement is on „Overall a good brand of water filter‟ and „Comes with an attractive
loan/installment‟. Other important product attribute perceptions such as efficacy, value for money
and company reputation have either remained the same or declined. Thus, the pilot has not
succeeded in establishing the perception that PureIt is an effective water treatment solution that
offers good value for money.
Thirdly, as is evident from the factors mentioned earlier, the target audience in Nagda had a more
positive perception of PureIt at endline than in Neemuch.
Table 14: Perceptions of PureIt
(mean score on a 3-point semantic differential scale)2
Neemuch Nagda
Baseline Endline Sig.1 Baseline Endline Sig.1
Overall, a good brand of water filter 2.88 2.68 * 2.68 2.85 *
Makes water safe 2.84 2.51 * 2.81 2.72
Is good value for money 2.60 2.43 2.63 2.63
Is from a reputed company 2.72 2.27 * 2.85 2.39 *
Has good after sales service 2.37 2.49 2.54 2.47
Is easy to use 2.70 2.53 2.85 2.75
Comes at an attractive price 2.52 2.38 2.29 2.50
Comes with an attractive loan/installment 2.16 2.32 2.22 2.62 *
Improves the quality of drinking water 2.81 2.53 * 2.71 2.82
Is a product in fashion 2.71 2.55 2.66 2.46
It offers better water quality than other
methods
2.78 2.47 * 2.65 2.71
Notes: 1Significance of difference between baseline and endline marked with * refers to p<0.05. 2 Attitudinal scale anchored by “Completely agree” (3) and “Disagree” (1)
These findings, interpreted in conjunction with the qualitative and quantitative research findings
regarding the target audience‟s recall of intervention activities suggests that:
In Nagda (the 100 percent cost-defrayment model), the pilot activities were not able to build
positive perceptions of the promoted solution (PureIt) in spite of a full run-time for the pilot. We
hypothesize that this could be because of two factors:
o Among the potential customers, the attractiveness of the cost-defrayment overshadowed
other messages (need for water treatment, effective methods), and
o Among the PureIt sales representatives, the ease of making a sale with cost-defrayment
could have led to lower emphasis on the core messages on need for water treatment and
ABT ASSOCIATES INC. 34
effective methods of water treatment, i.e. the sales representatives may not have needed to
persuade potential customers on the need for PureIt in order to make a sale.
Contrasting the findings on PureIt perceptions between Nagda and Neemuch could have provided
valuable insights into the effect of cost defrayment strategies on consumers and sales
representatives. For example, in Neemuch we expect the negative externality from cost-defrayment
to be less since a 50 percent cost-defrayment strategy was implemented here. However, due to the
short run-time for the pilot in Neemuch, we are unable to draw such inferences.
3.8. AWARENESS, PURCHASE AND USE OF PUREIT
There was a significant increase in spontaneous awareness, total awareness, purchase and current use
from baseline to endline in both pilot areas (Table 15). Awareness of PureIt had increased substantially
in both areas, reaching 56 percent in Neemuch and 74 percent in Nagda. At endline, the PureIt purchase
rate in Nagda (44 percent) was six times as high as in Neemuch, and current use (21 percent) was three
times as high. However, correct use remained low in both pilot areas, at less than 5 percent.
Table 15: Awareness, Purchase and Use of PureIt: Baseline-Endline Comparison
(percent of respondents)
Neemuch Nagda
Baseline Endline Sig.1 Baseline Endline Sig.1
Total Awareness 5.9 55.7 * 9.4 74.4 *
Spontaneous Awareness 4.0 27.4 * 5.0 52.4 *
Purchase 0.5 7.2 * 0.0 43.6 *
Current Use 0.4 3.7 * 0.0 21.2 *
Correct Use 0.0 1.8 0.0 4.3 * Note: 1Significance of difference between baseline and endline marked with * refers to p<0.05.
As shown in Figure 5, in both pilot areas there was approximately a 50 percent lapse rate (ratio of
current use to purchase). A comparison of the ratio of spontaneous awareness to total awareness in
Neemuch and Nagda is consistent with the findings from the surveys and qualitative research regarding
differences in the coverage and quality of product promotional activities. This ratio is lower in Neemuch
(48 percent) than in Nagda (70 percent), suggesting lower intensity of activity in Neemuch.
Furthermore, a comparison of the rate of conversion from awareness to purchase shows that the
conversion rate is much higher in Nagda (59 percent) than in Neemuch (13 percent).
ABT ASSOCIATES INC. 35
Figure 5: Awareness, Purchase and Curent Use of PureIt
(% of respondents)
Notes: SA = Spontaneous awareness; TA = Total awareness (including spontaneous and prompted awareness);
PU=purchased; CU= currently used
One of the reasons for this difference in conversion rates could be the difference in quality of promotional
activities. Other potential reasons for this difference are explored further in Figure 6 below. Since the
intervention was curtailed in Neemuch very early, we expected to find many respondents reporting that they
intended to buy but could not since the scheme was curtailed. However, only 8 percent of respondents in
Neemuch reported that they „intended to buy‟. Similarly in Nagda, 4 percent of respondents reported that
they „intended to buy‟, possibly because the scheme was stopped after six months when the ceiling on the
funds available for cost-defrayment was reached. Thus, the differences in uptake levels seen appear to be a
combination of the differences in the cost defrayment scheme (50 percent vs. 100 percent cost-defrayment)
and the quality and quantity of promotional activities in these pilots.
Figure 6: Potential market for Neemuch
27
5256
74
7
44
4
21
0
20
40
60
80
100
Neemuch Nagda
Spont Awareness
Total Awareness
Purchased
Currently used
SA/TA 48% PU/TA
13%
CU/PU 57%
SA/TA 70%
PU/TA 59%
CU/PU 48%
7
44
44
268
4
0
15
30
45
60
75
Neemuch Nagda
% r
esp
on
de
nts
(we
igh
ted
)
Purchased Not Aware Intend to Buy
59%
74%
ABT ASSOCIATES INC. 36
3.9. DETERMINANTS OF PURCHASE OF PUREIT
Those who were aware of PureIt but had not purchased it were asked why they did not do so. The
main reason reported for not purchasing PureIt was affordability (mentioned by about a third of
respondents who had not purchased the product), followed by the lack of availability of scheme (Table
16). Some reported an intention to buy in the near future and this was higher in Neemuch as mentioned
earlier.
Table 16: Reasons for Not Purchasing PureIt
(percent of respondents who knew of PureIt but did not buy it)
Reason for not buying Neemuch Nagda
No need for water treatment 1 1
Don‟t know where to get one 1 0
Cannot afford 39 33
Waiting for feedback on the product
Not sure about product/product quality 15 18
Scheme not available 19 21
Intend to buy 14 8
Affordability as the main reason for not purchasing PureIt is a surprising finding since the product was
being sold below the retail price. So, we examined the consistency of this finding through the association
between wealth status and purchase of PureIt (Figures 7a and 7b). If affordability is a barrier, we expect
to see a positive association between purchase and wealth status. The findings show that the uptake rate
is notably higher in higher wealth quintiles (particularly in Nagda where the pilot had a reasonable run
time) indicating affordability could be a barrier in spite of cost defrayment. Uptake (purchase rate) in the
SWP target group (middle three wealth quintiles) was 8 percent in Neemuch and 38 percent in Nagda;
while uptake in the richest quintile was 12 and 73 percent respectively. There was only one respondent
among the MFI clients who belonged to the poorest wealth quintile.
However, since very few of the MFI clients are from the highest wealth quintile, the distribution of
purchasers by wealth index shows that most of those who purchased PureIt (86 percent in Neemuch
and 67 percent in Nagda) were from the SWP target group. In Nagda, the units purchased by the SWP
target group were nearly all by households in the fourth quintile, with only 3 percent of units purchased
by the middle quintile and none in the second quintile. In Neemuch, out of 29 units that were sold in
total, seven were bought by households in the middle quintile, 18 were bought by households in the
fourth quintile, and the remaining four units were bought by households in the richest quintile.
ABT ASSOCIATES INC. 37
Figure 7a: Distribution of Purchasers by Wealth Status and Purchase Rates within Wealth Index
Quintiles at Endline, Neemuch10
Figure 7b: Distribution of Purchasers by Wealth Status and Purchase Rates within Wealth Index
Quintiles at Endline, Nagda11
10 There was only one respondent in the poorest wealth quintile in the total sample (in Neemuch). 11 There were no respondents in the sample in Nagda who belonged to the poorest wealth quintile.
ABT ASSOCIATES INC. 38
Additionally, we compared those who were aware of PureIt but had not purchased it with those who
purchased PureIt to identify the main psycho-social determinants of purchase and to describe the profile
of those who purchased PureIt. This was done through a segmentation analysis. It is the process of
dividing a heterogeneous population into homogenous groups based on the behavior of interest
(purchase of PureIt), and then identifying behavioral determinants and population characteristics that are
significantly different for those who purchased PureIt versus those who were aware of PureIt but did
not purchase it.
Table 17: Determinants of Purchase12
Determinants LOR Sign.
Brand Attitude:
PureIt is overall, a good brand of water filter 2.10 *
PureIt is good value for money 1.67 *
PureIt comes with an attractive loan/installment scheme 1.62 *
PureIt offers better water quality than other products and methods used to
treat water 1.97 *
Self-Efficacy: 2.61 *
I can take the required time to treat my water every time
I think it is easy to use filter to make water safe to drink
I am confident that I will be able to use filters to make water safe to drink
Wealth index: Low (ref)
Medium 1.76
High 3.22 *
Source of drinking water: Piped water to dwelling / plot
Public stand pipe 0.89
Tube well/bore well 4.56 *
Other sources 0.40
District: Neemuch (ref)
Nagda 6.77 *
HH with children under 5 years 0.51 *
Notes:
Significance: * refers to p<0.05.
LOR: Logistic odds ratios, derived through binary logistic regression.
ref = reference category
Backward stepwise procedure was used to develop a parsimonious model
Social norms, knowledge and attitudes are multi-item scales were individual items were measured on a semantic differential
scale anchored by “Completely Agree” (3) and “Disagree” (1)
12 This table compares the levels of determinants and population characteristics of those who purchased PureIt (N=212)
with those who were aware of PureIt but did not purchase it (N=334) at endline in project areas
ABT ASSOCIATES INC. 39
Logistic regression is used to find determinants of purchase. The results are presented as logistic odds
ratio (LOR) after controlling for population characteristics. Asterisks sign (*) indicates that the indicator
is significant at p<0.05 and is a determinant of purchase of PureIt. Table 17 presents those determinants
and population characteristics which are significantly associated with purchase and their LORs. The LOR
indicates the strength of relationship between the determinant and the desired behavior (purchase of
PureIt). An LOR of greater than one indicates that a particular determinant is likely to have a positive
effect on purchase of PureIt. Similarly, an LOR of less than one indicates that a determinant is likely to
have a negative effect on purchase of PureIt.
Table 17 indicates that those who thought that „PureIt is overall, a good brand of water filter‟, „PureIt is
good value for money‟, „PureIt comes with an attractive loan/installment scheme‟ and „PureIt offers
better water quality than other products and methods used to treat water‟ are more likely to purchase
PureIt than those who did not believe so. Those who had higher perceived self-efficacy (perceived ability
of treating water) were more likely to purchase. Therefore, positive brand attitude and perceived self-
efficacy are significant determinants to purchase of PureIt. Likelihood of purchase increased with wealth,
and households with tube/bore well were more likely to purchase PureIt than those who had their
water piped directly to their homes.
In summary, the segmentation analyses indicate that there were four factors that influenced purchase of
PureIt: wealth status, need for water treatment, perceptions of PureIt (efficacy and value for money) and